From 4c7cc51ca5ff888d4725f09a8d6f64ee71aa7e71 Mon Sep 17 00:00:00 2001 From: Ingo Oppermann Date: Thu, 31 Oct 2024 16:05:38 +0100 Subject: [PATCH] Update dependencies --- go.mod | 20 +- go.sum | 42 +- .../github.com/adhocore/gronx/.goreleaser.yml | 37 +- .../github.com/andybalholm/brotli/encoder.go | 9 + .../andybalholm/brotli/matchfinder/emitter.go | 11 - .../andybalholm/brotli/matchfinder/m4.go | 43 +- vendor/github.com/fatih/color/README.md | 23 +- vendor/github.com/fatih/color/color.go | 32 +- .../mimetype/internal/magic/binary.go | 2 + .../mimetype/internal/magic/ftyp.go | 43 +- .../mimetype/internal/magic/magic.go | 13 +- .../mimetype/internal/magic/ms_office.go | 45 +- .../mimetype/internal/magic/text_csv.go | 22 +- .../mimetype/internal/magic/zip.go | 89 +- .../mimetype/supported_mimes.md | 27 +- .../gabriel-vasile/mimetype/tree.go | 46 +- .../klauspost/compress/.goreleaser.yml | 6 +- .../github.com/klauspost/compress/README.md | 7 + .../klauspost/compress/s2/encode.go | 25 +- .../klauspost/compress/s2/encode_amd64.go | 201 +- .../klauspost/compress/s2/encode_go.go | 4 +- .../compress/s2/encodeblock_amd64.go | 44 +- .../klauspost/compress/s2/encodeblock_amd64.s | 21920 ++++++++-------- .../klauspost/compress/zstd/encoder.go | 26 +- .../klauspost/compress/zstd/zstd.go | 4 + vendor/github.com/minio/minio-go/v7/Makefile | 4 + vendor/github.com/minio/minio-go/v7/README.md | 2 +- .../minio-go/v7/api-put-object-streaming.go | 2 +- .../minio/minio-go/v7/api-put-object.go | 2 + vendor/github.com/minio/minio-go/v7/api.go | 18 +- .../minio/minio-go/v7/functional_tests.go | 14 +- .../v7/pkg/credentials/sts_web_identity.go | 44 +- .../minio-go/v7/pkg/lifecycle/lifecycle.go | 26 + .../minio/minio-go/v7/pkg/tags/tags.go | 2 +- vendor/github.com/swaggo/swag/Dockerfile | 10 +- vendor/github.com/swaggo/swag/Makefile | 13 +- vendor/github.com/swaggo/swag/README.md | 31 +- vendor/github.com/swaggo/swag/README_pt.md | 2 +- vendor/github.com/swaggo/swag/README_zh-CN.md | 3 +- vendor/github.com/swaggo/swag/field_parser.go | 54 +- vendor/github.com/swaggo/swag/generics.go | 20 +- vendor/github.com/swaggo/swag/operation.go | 12 +- vendor/github.com/swaggo/swag/packages.go | 63 +- vendor/github.com/swaggo/swag/parser.go | 245 +- vendor/github.com/swaggo/swag/types.go | 46 +- vendor/github.com/swaggo/swag/version.go | 2 +- .../validator/rules/fields_on_correct_type.go | 39 +- .../validator/rules/known_argument_names.go | 70 +- .../v2/validator/rules/known_type_names.go | 88 +- .../validator/rules/values_of_correct_type.go | 183 +- .../gqlparser/v2/validator/validator.go | 11 + vendor/modules.txt | 24 +- 52 files changed, 12229 insertions(+), 11542 deletions(-) diff --git a/go.mod b/go.mod index 47f22d5c..bd4bed8a 100644 --- a/go.mod +++ b/go.mod @@ -7,8 +7,8 @@ toolchain go1.23.1 require ( github.com/99designs/gqlgen v0.17.55 github.com/Masterminds/semver/v3 v3.3.0 - github.com/adhocore/gronx v1.19.1 - github.com/andybalholm/brotli v1.1.0 + github.com/adhocore/gronx v1.19.3 + github.com/andybalholm/brotli v1.1.1 github.com/atrox/haikunatorgo/v2 v2.0.1 github.com/caddyserver/certmagic v0.21.4 github.com/datarhei/gosrt v0.7.0 @@ -26,21 +26,21 @@ require ( github.com/hashicorp/raft-boltdb/v2 v2.3.0 github.com/invopop/jsonschema v0.4.0 github.com/joho/godotenv v1.5.1 - github.com/klauspost/compress v1.17.10 + github.com/klauspost/compress v1.17.11 github.com/klauspost/cpuid/v2 v2.2.8 github.com/labstack/echo/v4 v4.12.0 github.com/lestrrat-go/strftime v1.1.0 github.com/lithammer/shortuuid/v4 v4.0.0 github.com/mattn/go-isatty v0.0.20 - github.com/minio/minio-go/v7 v7.0.77 - github.com/prometheus/client_golang v1.20.4 + github.com/minio/minio-go/v7 v7.0.80 + github.com/prometheus/client_golang v1.20.5 github.com/puzpuzpuz/xsync/v3 v3.4.0 github.com/shirou/gopsutil/v3 v3.24.5 github.com/stretchr/testify v1.9.0 github.com/swaggo/echo-swagger v1.4.1 - github.com/swaggo/swag v1.16.3 + github.com/swaggo/swag v1.16.4 github.com/tklauser/go-sysconf v0.3.14 - github.com/vektah/gqlparser/v2 v2.5.17 + github.com/vektah/gqlparser/v2 v2.5.18 github.com/xeipuuv/gojsonschema v1.2.0 go.etcd.io/bbolt v1.3.11 go.uber.org/automaxprocs v1.6.0 @@ -62,8 +62,8 @@ require ( github.com/davecgh/go-spew v1.1.1 // indirect github.com/dolthub/maphash v0.1.0 // indirect github.com/dustin/go-humanize v1.0.1 // indirect - github.com/fatih/color v1.17.0 // indirect - github.com/gabriel-vasile/mimetype v1.4.5 // indirect + github.com/fatih/color v1.18.0 // indirect + github.com/gabriel-vasile/mimetype v1.4.6 // indirect github.com/ghodss/yaml v1.0.0 // indirect github.com/go-ini/ini v1.67.0 // indirect github.com/go-ole/go-ole v1.3.0 // indirect @@ -96,7 +96,7 @@ require ( github.com/pmezard/go-difflib v1.0.1-0.20181226105442-5d4384ee4fb2 // indirect github.com/power-devops/perfstat v0.0.0-20240221224432-82ca36839d55 // indirect github.com/prometheus/client_model v0.6.1 // indirect - github.com/prometheus/common v0.60.0 // indirect + github.com/prometheus/common v0.60.1 // indirect github.com/prometheus/procfs v0.15.1 // indirect github.com/rs/xid v1.6.0 // indirect github.com/russross/blackfriday/v2 v2.1.0 // indirect diff --git a/go.sum b/go.sum index 95462ea3..002f356e 100644 --- a/go.sum +++ b/go.sum @@ -7,8 +7,8 @@ github.com/Masterminds/semver/v3 v3.3.0 h1:B8LGeaivUe71a5qox1ICM/JLl0NqZSW5CHyL+ github.com/Masterminds/semver/v3 v3.3.0/go.mod h1:4V+yj/TJE1HU9XfppCwVMZq3I84lprf4nC11bSS5beM= github.com/PuerkitoBio/goquery v1.9.3 h1:mpJr/ikUA9/GNJB/DBZcGeFDXUtosHRyRrwh7KGdTG0= github.com/PuerkitoBio/goquery v1.9.3/go.mod h1:1ndLHPdTz+DyQPICCWYlYQMPl0oXZj0G6D4LCYA6u4U= -github.com/adhocore/gronx v1.19.1 h1:S4c3uVp5jPjnk00De0lslyTenGJ4nA3Ydbkj1SbdPVc= -github.com/adhocore/gronx v1.19.1/go.mod h1:7oUY1WAU8rEJWmAxXR2DN0JaO4gi9khSgKjiRypqteg= +github.com/adhocore/gronx v1.19.3 h1:BNl3pGgiKy12Tt1dNIhzEGL2l+KikeV6NN3A/B2AJLg= +github.com/adhocore/gronx v1.19.3/go.mod h1:7oUY1WAU8rEJWmAxXR2DN0JaO4gi9khSgKjiRypqteg= github.com/agnivade/levenshtein v1.2.0 h1:U9L4IOT0Y3i0TIlUIDJ7rVUziKi/zPbrJGaFrtYH3SY= github.com/agnivade/levenshtein v1.2.0/go.mod h1:QVVI16kDrtSuwcpd0p1+xMC6Z/VfhtCyDIjcwga4/DU= github.com/alecthomas/template v0.0.0-20160405071501-a0175ee3bccc/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc= @@ -17,8 +17,8 @@ github.com/alecthomas/units v0.0.0-20151022065526-2efee857e7cf/go.mod h1:ybxpYRF github.com/alecthomas/units v0.0.0-20190717042225-c3de453c63f4/go.mod h1:ybxpYRFXyAe+OPACYpWeL0wqObRcbAqCMya13uyzqw0= github.com/andreyvit/diff v0.0.0-20170406064948-c7f18ee00883 h1:bvNMNQO63//z+xNgfBlViaCIJKLlCJ6/fmUseuG0wVQ= github.com/andreyvit/diff v0.0.0-20170406064948-c7f18ee00883/go.mod h1:rCTlJbsFo29Kk6CurOXKm700vrz8f0KW0JNfpkRJY/8= -github.com/andybalholm/brotli v1.1.0 h1:eLKJA0d02Lf0mVpIDgYnqXcUn0GqVmEFny3VuID1U3M= -github.com/andybalholm/brotli v1.1.0/go.mod h1:sms7XGricyQI9K10gOSf56VKKWS4oLer58Q+mhRPtnY= +github.com/andybalholm/brotli v1.1.1 h1:PR2pgnyFznKEugtsUo0xLdDop5SKXd5Qf5ysW+7XdTA= +github.com/andybalholm/brotli v1.1.1/go.mod h1:05ib4cKhjx3OQYUY22hTVd34Bc8upXjOLL2rKwwZBoA= github.com/andybalholm/cascadia v1.3.2 h1:3Xi6Dw5lHF15JtdcmAHD3i1+T8plmv7BQ/nsViSLyss= github.com/andybalholm/cascadia v1.3.2/go.mod h1:7gtRlve5FxPPgIgX36uWBX58OdBsSS6lUvCFb+h7KvU= github.com/arbovm/levenshtein v0.0.0-20160628152529-48b4e1c0c4d0 h1:jfIu9sQUG6Ig+0+Ap1h4unLjW6YQJpKZVmUzxsD4E/Q= @@ -63,12 +63,12 @@ github.com/dustin/go-humanize v1.0.0/go.mod h1:HtrtbFcZ19U5GC7JDqmcUSB87Iq5E25Kn github.com/dustin/go-humanize v1.0.1 h1:GzkhY7T5VNhEkwH0PVJgjz+fX1rhBrR7pRT3mDkpeCY= github.com/dustin/go-humanize v1.0.1/go.mod h1:Mu1zIs6XwVuF/gI1OepvI0qD18qycQx+mFykh5fBlto= github.com/fatih/color v1.13.0/go.mod h1:kLAiJbzzSOZDVNGyDpeOxJ47H46qBXwg5ILebYFFOfk= -github.com/fatih/color v1.17.0 h1:GlRw1BRJxkpqUCBKzKOw098ed57fEsKeNjpTe3cSjK4= -github.com/fatih/color v1.17.0/go.mod h1:YZ7TlrGPkiz6ku9fK3TLD/pl3CpsiFyu8N92HLgmosI= +github.com/fatih/color v1.18.0 h1:S8gINlzdQ840/4pfAwic/ZE0djQEH3wM94VfqLTZcOM= +github.com/fatih/color v1.18.0/go.mod h1:4FelSpRwEGDpQ12mAdzqdOukCy4u8WUtOY6lkT/6HfU= github.com/fujiwara/shapeio v1.0.0 h1:xG5D9oNqCSUUbryZ/jQV3cqe1v2suEjwPIcEg1gKM8M= github.com/fujiwara/shapeio v1.0.0/go.mod h1:LmEmu6L/8jetyj1oewewFb7bZCNRwE7wLCUNzDLaLVA= -github.com/gabriel-vasile/mimetype v1.4.5 h1:J7wGKdGu33ocBOhGy0z653k/lFKLFDPJMG8Gql0kxn4= -github.com/gabriel-vasile/mimetype v1.4.5/go.mod h1:ibHel+/kbxn9x2407k1izTA1S81ku1z/DlgOW2QE0M4= +github.com/gabriel-vasile/mimetype v1.4.6 h1:3+PzJTKLkvgjeTbts6msPJt4DixhT4YtFNf1gtGe3zc= +github.com/gabriel-vasile/mimetype v1.4.6/go.mod h1:JX1qVKqZd40hUPpAfiNTe0Sne7hdfKSbOqqmkq8GCXc= github.com/ghodss/yaml v1.0.0 h1:wQHKEahhL6wmXdzwWG11gIVCkOv05bNOh+Rxn0yngAk= github.com/ghodss/yaml v1.0.0/go.mod h1:4dBDuWmgqj2HViK6kFavaiC9ZROes6MMH2rRYeMEF04= github.com/go-ini/ini v1.67.0 h1:z6ZrTEZqSWOTyH2FlglNbNgARyHG8oLW9gMELqKr06A= @@ -159,8 +159,8 @@ github.com/josharian/intern v1.0.0/go.mod h1:5DoeVV0s6jJacbCEi61lwdGj/aVlrQvzHFF github.com/json-iterator/go v1.1.6/go.mod h1:+SdeFBvtyEkXs7REEP0seUULqWtbJapLOCVDaaPEHmU= github.com/json-iterator/go v1.1.9/go.mod h1:KdQUCv79m/52Kvf8AW2vK1V8akMuk1QjK/uOdHXbAo4= github.com/julienschmidt/httprouter v1.2.0/go.mod h1:SYymIcj16QtmaHHD7aYtjjsJG7VTCxuUUipMqKk8s4w= -github.com/klauspost/compress v1.17.10 h1:oXAz+Vh0PMUvJczoi+flxpnBEPxoER1IaAnU/NMPtT0= -github.com/klauspost/compress v1.17.10/go.mod h1:pMDklpSncoRMuLFrf1W9Ss9KT+0rH90U12bZKk7uwG0= +github.com/klauspost/compress v1.17.11 h1:In6xLpyWOi1+C7tXUUWv2ot1QvBjxevKAaI6IXrJmUc= +github.com/klauspost/compress v1.17.11/go.mod h1:pMDklpSncoRMuLFrf1W9Ss9KT+0rH90U12bZKk7uwG0= github.com/klauspost/cpuid/v2 v2.0.1/go.mod h1:FInQzS24/EEf25PyTYn52gqo7WaD8xa0213Md/qVLRg= github.com/klauspost/cpuid/v2 v2.2.8 h1:+StwCXwm9PdpiEkPyzBXIy+M9KUb4ODm0Zarf1kS5BM= github.com/klauspost/cpuid/v2 v2.2.8/go.mod h1:Lcz8mBdAVJIBVzewtcLocK12l3Y+JytZYpaMropDUws= @@ -209,8 +209,8 @@ github.com/miekg/dns v1.1.62 h1:cN8OuEF1/x5Rq6Np+h1epln8OiyPWV+lROx9LxcGgIQ= github.com/miekg/dns v1.1.62/go.mod h1:mvDlcItzm+br7MToIKqkglaGhlFMHJ9DTNNWONWXbNQ= github.com/minio/md5-simd v1.1.2 h1:Gdi1DZK69+ZVMoNHRXJyNcxrMA4dSxoYHZSQbirFg34= github.com/minio/md5-simd v1.1.2/go.mod h1:MzdKDxYpY2BT9XQFocsiZf/NKVtR7nkE4RoEpN+20RM= -github.com/minio/minio-go/v7 v7.0.77 h1:GaGghJRg9nwDVlNbwYjSDJT1rqltQkBFDsypWX1v3Bw= -github.com/minio/minio-go/v7 v7.0.77/go.mod h1:AVM3IUN6WwKzmwBxVdjzhH8xq+f57JSbbvzqvUzR6eg= +github.com/minio/minio-go/v7 v7.0.80 h1:2mdUHXEykRdY/BigLt3Iuu1otL0JTogT0Nmltg0wujk= +github.com/minio/minio-go/v7 v7.0.80/go.mod h1:84gmIilaX4zcvAWWzJ5Z1WI5axN+hAbM5w25xf8xvC0= github.com/mitchellh/mapstructure v1.5.0 h1:jeMsZIYE/09sWLaz43PL7Gy6RuMjD2eJVyuac5Z2hdY= github.com/mitchellh/mapstructure v1.5.0/go.mod h1:bFUtVrKA4DC2yAKiSyO/QUcy7e+RRV2QTWOzhPopBRo= github.com/modern-go/concurrent v0.0.0-20180228061459-e0a39a4cb421/go.mod h1:6dJC0mAP4ikYIbvyc7fijjWJddQyLn8Ig3JB5CqoB9Q= @@ -234,8 +234,8 @@ github.com/prashantv/gostub v1.1.0/go.mod h1:A5zLQHz7ieHGG7is6LLXLz7I8+3LZzsrV0P github.com/prometheus/client_golang v0.9.1/go.mod h1:7SWBe2y4D6OKWSNQJUaRYU/AaXPKyh/dDVn+NZz0KFw= github.com/prometheus/client_golang v1.0.0/go.mod h1:db9x61etRT2tGnBNRi70OPL5FsnadC4Ky3P0J6CfImo= github.com/prometheus/client_golang v1.4.0/go.mod h1:e9GMxYsXl05ICDXkRhurwBS4Q3OK1iX/F2sw+iXX5zU= -github.com/prometheus/client_golang v1.20.4 h1:Tgh3Yr67PaOv/uTqloMsCEdeuFTatm5zIq5+qNN23vI= -github.com/prometheus/client_golang v1.20.4/go.mod h1:PIEt8X02hGcP8JWbeHyeZ53Y/jReSnHgO035n//V5WE= +github.com/prometheus/client_golang v1.20.5 h1:cxppBPuYhUnsO6yo/aoRol4L7q7UFfdm+bR9r+8l63Y= +github.com/prometheus/client_golang v1.20.5/go.mod h1:PIEt8X02hGcP8JWbeHyeZ53Y/jReSnHgO035n//V5WE= github.com/prometheus/client_model v0.0.0-20180712105110-5c3871d89910/go.mod h1:MbSGuTsp3dbXC40dX6PRTWyKYBIrTGTE9sqQNg2J8bo= github.com/prometheus/client_model v0.0.0-20190129233127-fd36f4220a90/go.mod h1:xMI15A0UPsDsEKsMN9yxemIoYk6Tm2C1GtYGdfGttqA= github.com/prometheus/client_model v0.2.0/go.mod h1:xMI15A0UPsDsEKsMN9yxemIoYk6Tm2C1GtYGdfGttqA= @@ -243,8 +243,8 @@ github.com/prometheus/client_model v0.6.1 h1:ZKSh/rekM+n3CeS952MLRAdFwIKqeY8b62p github.com/prometheus/client_model v0.6.1/go.mod h1:OrxVMOVHjw3lKMa8+x6HeMGkHMQyHDk9E3jmP2AmGiY= github.com/prometheus/common v0.4.1/go.mod h1:TNfzLD0ON7rHzMJeJkieUDPYmFC7Snx/y86RQel1bk4= github.com/prometheus/common v0.9.1/go.mod h1:yhUN8i9wzaXS3w1O07YhxHEBxD+W35wd8bs7vj7HSQ4= -github.com/prometheus/common v0.60.0 h1:+V9PAREWNvJMAuJ1x1BaWl9dewMW4YrHZQbx0sJNllA= -github.com/prometheus/common v0.60.0/go.mod h1:h0LYf1R1deLSKtD4Vdg8gy4RuOvENW2J/h19V5NADQw= +github.com/prometheus/common v0.60.1 h1:FUas6GcOw66yB/73KC+BOZoFJmbo/1pojoILArPAaSc= +github.com/prometheus/common v0.60.1/go.mod h1:h0LYf1R1deLSKtD4Vdg8gy4RuOvENW2J/h19V5NADQw= github.com/prometheus/procfs v0.0.0-20181005140218-185b4288413d/go.mod h1:c3At6R/oaqEKCNdg8wHV1ftS6bRYblBhIjjI8uT2IGk= github.com/prometheus/procfs v0.0.2/go.mod h1:TjEm7ze935MbeOT/UhFTIMYKhuLP4wbCsTZCD3I8kEA= github.com/prometheus/procfs v0.0.8/go.mod h1:7Qr8sr6344vo1JqZ6HhLceV9o3AJ1Ff+GxbHq6oeK9A= @@ -283,8 +283,8 @@ github.com/swaggo/echo-swagger v1.4.1 h1:Yf0uPaJWp1uRtDloZALyLnvdBeoEL5Kc7DtnjzO github.com/swaggo/echo-swagger v1.4.1/go.mod h1:C8bSi+9yH2FLZsnhqMZLIZddpUxZdBYuNHbtaS1Hljc= github.com/swaggo/files/v2 v2.0.1 h1:XCVJO/i/VosCDsJu1YLpdejGsGnBE9deRMpjN4pJLHk= github.com/swaggo/files/v2 v2.0.1/go.mod h1:24kk2Y9NYEJ5lHuCra6iVwkMjIekMCaFq/0JQj66kyM= -github.com/swaggo/swag v1.16.3 h1:PnCYjPCah8FK4I26l2F/KQ4yz3sILcVUN3cTlBFA9Pg= -github.com/swaggo/swag v1.16.3/go.mod h1:DImHIuOFXKpMFAQjcC7FG4m3Dg4+QuUgUzJmKjI/gRk= +github.com/swaggo/swag v1.16.4 h1:clWJtd9LStiG3VeijiCfOVODP6VpHtKdQy9ELFG3s1A= +github.com/swaggo/swag v1.16.4/go.mod h1:VBsHJRsDvfYvqoiMKnsdwhNV9LEMHgEDZcyVYX0sxPg= github.com/tklauser/go-sysconf v0.3.14 h1:g5vzr9iPFFz24v2KZXs/pvpvh8/V9Fw6vQK5ZZb78yU= github.com/tklauser/go-sysconf v0.3.14/go.mod h1:1ym4lWMLUOhuBOPGtRcJm7tEGX4SCYNEEEtghGG/8uY= github.com/tklauser/numcpus v0.9.0 h1:lmyCHtANi8aRUgkckBgoDk1nHCux3n2cgkJLXdQGPDo= @@ -296,8 +296,8 @@ github.com/valyala/bytebufferpool v1.0.0 h1:GqA5TC/0021Y/b9FG4Oi9Mr3q7XYx6Kllzaw github.com/valyala/bytebufferpool v1.0.0/go.mod h1:6bBcMArwyJ5K/AmCkWv1jt77kVWyCJ6HpOuEn7z0Csc= github.com/valyala/fasttemplate v1.2.2 h1:lxLXG0uE3Qnshl9QyaK6XJxMXlQZELvChBOCmQD0Loo= github.com/valyala/fasttemplate v1.2.2/go.mod h1:KHLXt3tVN2HBp8eijSv/kGJopbvo7S+qRAEEKiv+SiQ= -github.com/vektah/gqlparser/v2 v2.5.17 h1:9At7WblLV7/36nulgekUgIaqHZWn5hxqluxrxGUhOmI= -github.com/vektah/gqlparser/v2 v2.5.17/go.mod h1:1lz1OeCqgQbQepsGxPVywrjdBHW2T08PUS3pJqepRww= +github.com/vektah/gqlparser/v2 v2.5.18 h1:zSND3GtutylAQ1JpWnTHcqtaRZjl+y3NROeW8vuNo6Y= +github.com/vektah/gqlparser/v2 v2.5.18/go.mod h1:6HLzf7JKv9Fi3APymudztFQNmLXR5qJeEo6BOFcXVfc= github.com/xeipuuv/gojsonpointer v0.0.0-20180127040702-4e3ac2762d5f/go.mod h1:N2zxlSyiKSe5eX1tZViRH5QA0qijqEDrYZiPEAiq3wU= github.com/xeipuuv/gojsonpointer v0.0.0-20190905194746-02993c407bfb h1:zGWFAtiMcyryUHoUjUJX0/lt1H2+i2Ka2n+D3DImSNo= github.com/xeipuuv/gojsonpointer v0.0.0-20190905194746-02993c407bfb/go.mod h1:N2zxlSyiKSe5eX1tZViRH5QA0qijqEDrYZiPEAiq3wU= @@ -307,6 +307,8 @@ github.com/xeipuuv/gojsonschema v1.2.0 h1:LhYJRs+L4fBtjZUfuSZIKGeVu0QRy8e5Xi7D17 github.com/xeipuuv/gojsonschema v1.2.0/go.mod h1:anYRn/JVcOK2ZgGU+IjEV4nwlhoK5sQluxsYJ78Id3Y= github.com/xrash/smetrics v0.0.0-20240521201337-686a1a2994c1 h1:gEOO8jv9F4OT7lGCjxCBTO/36wtF6j2nSip77qHd4x4= github.com/xrash/smetrics v0.0.0-20240521201337-686a1a2994c1/go.mod h1:Ohn+xnUBiLI6FVj/9LpzZWtj1/D6lUovWYBkxHVV3aM= +github.com/xyproto/randomstring v1.0.5 h1:YtlWPoRdgMu3NZtP45drfy1GKoojuR7hmRcnhZqKjWU= +github.com/xyproto/randomstring v1.0.5/go.mod h1:rgmS5DeNXLivK7YprL0pY+lTuhNQW3iGxZ18UQApw/E= github.com/yusufpapurcu/wmi v1.2.4 h1:zFUKzehAFReQwLys1b/iSMl+JQGSCSjtVqQn9bBrPo0= github.com/yusufpapurcu/wmi v1.2.4/go.mod h1:SBZ9tNy3G9/m5Oi98Zks0QjeHVDvuK0qfxQmPyzfmi0= github.com/zeebo/assert v1.1.0 h1:hU1L1vLTHsnO8x8c9KAR5GmM5QscxHg5RNU5z5qbUWY= diff --git a/vendor/github.com/adhocore/gronx/.goreleaser.yml b/vendor/github.com/adhocore/gronx/.goreleaser.yml index b52f6d99..242ae75a 100644 --- a/vendor/github.com/adhocore/gronx/.goreleaser.yml +++ b/vendor/github.com/adhocore/gronx/.goreleaser.yml @@ -1,3 +1,8 @@ + +# yaml-language-server: $schema=https://goreleaser.com/static/schema.json + +version: 2 + project_name: tasker release: @@ -11,41 +16,37 @@ before: - go mod tidy builds: - - <<: &build_defaults - binary: bin/tasker - main: ./cmd/tasker - ldflags: - - -X main.Version={{.Version}} - env: - - CGO_ENABLED=0 - id: macOS + - id: macOS + binary: bin/tasker + main: ./cmd/tasker + ldflags: + - -X main.Version={{.Version}} + env: + - CGO_ENABLED=0 goos: [darwin] goarch: [amd64, arm64] - - <<: *build_defaults - id: linux + - id: linux + main: ./cmd/tasker goos: [linux] - goarch: [386, arm, amd64, arm64] + goarch: ["386", arm, amd64, arm64] - - <<: *build_defaults - id: windows + - id: windows + main: ./cmd/tasker goos: [windows] goarch: [amd64] archives: - id: nix builds: [macOS, linux] - <<: &archive_defaults - name_template: "{{ .ProjectName }}_{{ .Version }}_{{ .Os }}_{{ .Arch }}{{ if .Arm }}v{{ .Arm }}{{ end }}" + name_template: "{{ .ProjectName }}_{{ .Version }}_{{ .Os }}_{{ .Arch }}{{ if .Arm }}v{{ .Arm }}{{ end }}" wrap_in_directory: true - rlcp: true format: tar.gz files: - LICENSE - id: windows builds: [windows] - <<: *archive_defaults wrap_in_directory: false format: zip files: @@ -56,7 +57,7 @@ checksum: algorithm: sha256 changelog: - skip: true + disable: true use: github sort: desc filters: diff --git a/vendor/github.com/andybalholm/brotli/encoder.go b/vendor/github.com/andybalholm/brotli/encoder.go index 650d1e42..19283825 100644 --- a/vendor/github.com/andybalholm/brotli/encoder.go +++ b/vendor/github.com/andybalholm/brotli/encoder.go @@ -21,6 +21,15 @@ func (e *Encoder) Encode(dst []byte, src []byte, matches []matchfinder.Match, la e.wroteHeader = true } + if len(src) == 0 { + if lastBlock { + e.bw.writeBits(2, 3) // islast + isempty + e.bw.jumpToByteBoundary() + return e.bw.dst + } + return dst + } + var literalHisto [256]uint32 var commandHisto [704]uint32 var distanceHisto [64]uint32 diff --git a/vendor/github.com/andybalholm/brotli/matchfinder/emitter.go b/vendor/github.com/andybalholm/brotli/matchfinder/emitter.go index 37ed8e13..507d1cae 100644 --- a/vendor/github.com/andybalholm/brotli/matchfinder/emitter.go +++ b/vendor/github.com/andybalholm/brotli/matchfinder/emitter.go @@ -32,14 +32,3 @@ func (e *matchEmitter) emit(m absoluteMatch) { }) e.NextEmit = m.End } - -// trim shortens m if it extends past maxEnd. Then if the length is at least -// minLength, the match is emitted. -func (e *matchEmitter) trim(m absoluteMatch, maxEnd int, minLength int) { - if m.End > maxEnd { - m.End = maxEnd - } - if m.End-m.Start >= minLength { - e.emit(m) - } -} diff --git a/vendor/github.com/andybalholm/brotli/matchfinder/m4.go b/vendor/github.com/andybalholm/brotli/matchfinder/m4.go index 5b2acba2..81894725 100644 --- a/vendor/github.com/andybalholm/brotli/matchfinder/m4.go +++ b/vendor/github.com/andybalholm/brotli/matchfinder/m4.go @@ -56,7 +56,7 @@ func (q *M4) Reset() { } func (q *M4) score(m absoluteMatch) int { - return (m.End-m.Start)*256 + bits.LeadingZeros32(uint32(m.Start-m.Match))*q.DistanceBitCost + return (m.End-m.Start)*256 + (bits.LeadingZeros32(uint32(m.Start-m.Match))-32)*q.DistanceBitCost } func (q *M4) FindMatches(dst []Match, src []byte) []Match { @@ -112,7 +112,12 @@ func (q *M4) FindMatches(dst []Match, src []byte) []Match { // We have found some matches, and we're far enough along that we probably // won't find overlapping matches, so we might as well emit them. if matches[1] != (absoluteMatch{}) { - e.trim(matches[1], matches[0].Start, q.MinLength) + if matches[1].End > matches[0].Start { + matches[1].End = matches[0].Start + } + if matches[1].End-matches[1].Start >= q.MinLength && q.score(matches[1]) > 0 { + e.emit(matches[1]) + } } e.emit(matches[0]) matches = [3]absoluteMatch{} @@ -139,12 +144,10 @@ func (q *M4) FindMatches(dst []Match, src []byte) []Match { // Look for a match. var currentMatch absoluteMatch - if i-candidate != matches[0].Start-matches[0].Match { - if binary.LittleEndian.Uint32(src[candidate:]) == binary.LittleEndian.Uint32(src[i:]) { - m := extendMatch2(src, i, candidate, e.NextEmit) - if m.End-m.Start > q.MinLength { - currentMatch = m - } + if binary.LittleEndian.Uint32(src[candidate:]) == binary.LittleEndian.Uint32(src[i:]) { + m := extendMatch2(src, i, candidate, e.NextEmit) + if m.End-m.Start > q.MinLength && q.score(m) > 0 { + currentMatch = m } } @@ -157,12 +160,10 @@ func (q *M4) FindMatches(dst []Match, src []byte) []Match { if candidate <= 0 || i-candidate > q.MaxDistance { break } - if i-candidate != matches[0].Start-matches[0].Match { - if binary.LittleEndian.Uint32(src[candidate:]) == binary.LittleEndian.Uint32(src[i:]) { - m := extendMatch2(src, i, candidate, e.NextEmit) - if m.End-m.Start > q.MinLength && q.score(m) > q.score(currentMatch) { - currentMatch = m - } + if binary.LittleEndian.Uint32(src[candidate:]) == binary.LittleEndian.Uint32(src[i:]) { + m := extendMatch2(src, i, candidate, e.NextEmit) + if m.End-m.Start > q.MinLength && q.score(m) > q.score(currentMatch) { + currentMatch = m } } } @@ -217,14 +218,24 @@ func (q *M4) FindMatches(dst []Match, src []byte) []Match { default: // Emit the first match, shortening it if necessary to avoid overlap with the second. - e.trim(matches[2], matches[1].Start, q.MinLength) + if matches[2].End > matches[1].Start { + matches[2].End = matches[1].Start + } + if matches[2].End-matches[2].Start >= q.MinLength && q.score(matches[2]) > 0 { + e.emit(matches[2]) + } matches[2] = absoluteMatch{} } } // We've found all the matches now; emit the remaining ones. if matches[1] != (absoluteMatch{}) { - e.trim(matches[1], matches[0].Start, q.MinLength) + if matches[1].End > matches[0].Start { + matches[1].End = matches[0].Start + } + if matches[1].End-matches[1].Start >= q.MinLength && q.score(matches[1]) > 0 { + e.emit(matches[1]) + } } if matches[0] != (absoluteMatch{}) { e.emit(matches[0]) diff --git a/vendor/github.com/fatih/color/README.md b/vendor/github.com/fatih/color/README.md index be82827c..d135bfe0 100644 --- a/vendor/github.com/fatih/color/README.md +++ b/vendor/github.com/fatih/color/README.md @@ -9,7 +9,7 @@ suits you. ## Install -```bash +``` go get github.com/fatih/color ``` @@ -30,6 +30,18 @@ color.Magenta("And many others ..") ``` +### RGB colors + +If your terminal supports 24-bit colors, you can use RGB color codes. + +```go +color.RGB(255, 128, 0).Println("foreground orange") +color.RGB(230, 42, 42).Println("foreground red") + +color.BgRGB(255, 128, 0).Println("background orange") +color.BgRGB(230, 42, 42).Println("background red") +``` + ### Mix and reuse colors ```go @@ -49,6 +61,11 @@ boldRed.Println("This will print text in bold red.") whiteBackground := red.Add(color.BgWhite) whiteBackground.Println("Red text with white background.") + +// Mix with RGB color codes +color.RGB(255, 128, 0).AddBgRGB(0, 0, 0).Println("orange with black background") + +color.BgRGB(255, 128, 0).AddRGB(255, 255, 255).Println("orange background with white foreground") ``` ### Use your own output (io.Writer) @@ -161,10 +178,6 @@ c.Println("This prints again cyan...") To output color in GitHub Actions (or other CI systems that support ANSI colors), make sure to set `color.NoColor = false` so that it bypasses the check for non-tty output streams. -## Todo - -* Save/Return previous values -* Evaluate fmt.Formatter interface ## Credits diff --git a/vendor/github.com/fatih/color/color.go b/vendor/github.com/fatih/color/color.go index 81094e87..ee39b408 100644 --- a/vendor/github.com/fatih/color/color.go +++ b/vendor/github.com/fatih/color/color.go @@ -98,6 +98,9 @@ const ( FgMagenta FgCyan FgWhite + + // used internally for 256 and 24-bit coloring + foreground ) // Foreground Hi-Intensity text colors @@ -122,6 +125,9 @@ const ( BgMagenta BgCyan BgWhite + + // used internally for 256 and 24-bit coloring + background ) // Background Hi-Intensity text colors @@ -150,6 +156,30 @@ func New(value ...Attribute) *Color { return c } +// RGB returns a new foreground color in 24-bit RGB. +func RGB(r, g, b int) *Color { + return New(foreground, 2, Attribute(r), Attribute(g), Attribute(b)) +} + +// BgRGB returns a new background color in 24-bit RGB. +func BgRGB(r, g, b int) *Color { + return New(background, 2, Attribute(r), Attribute(g), Attribute(b)) +} + +// AddRGB is used to chain foreground RGB SGR parameters. Use as many as parameters to combine +// and create custom color objects. Example: .Add(34, 0, 12).Add(255, 128, 0). +func (c *Color) AddRGB(r, g, b int) *Color { + c.params = append(c.params, foreground, 2, Attribute(r), Attribute(g), Attribute(b)) + return c +} + +// AddRGB is used to chain background RGB SGR parameters. Use as many as parameters to combine +// and create custom color objects. Example: .Add(34, 0, 12).Add(255, 128, 0). +func (c *Color) AddBgRGB(r, g, b int) *Color { + c.params = append(c.params, background, 2, Attribute(r), Attribute(g), Attribute(b)) + return c +} + // Set sets the given parameters immediately. It will change the color of // output with the given SGR parameters until color.Unset() is called. func Set(p ...Attribute) *Color { @@ -401,7 +431,7 @@ func (c *Color) format() string { func (c *Color) unformat() string { //return fmt.Sprintf("%s[%dm", escape, Reset) - //for each element in sequence let's use the speficic reset escape, ou the generic one if not found + //for each element in sequence let's use the specific reset escape, or the generic one if not found format := make([]string, len(c.params)) for i, v := range c.params { format[i] = strconv.Itoa(int(Reset)) diff --git a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/binary.go b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/binary.go index f1e94498..96040e6c 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/binary.go +++ b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/binary.go @@ -21,6 +21,8 @@ var ( SWF = prefix([]byte("CWS"), []byte("FWS"), []byte("ZWS")) // Torrent has bencoded text in the beginning. Torrent = prefix([]byte("d8:announce")) + // PAR1 matches a parquet file. + Par1 = prefix([]byte{0x50, 0x41, 0x52, 0x31}) ) // Java bytecode and Mach-O binaries share the same magic number. diff --git a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ftyp.go b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ftyp.go index 6575b4ae..ac727139 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ftyp.go +++ b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ftyp.go @@ -1,22 +1,14 @@ package magic -import "bytes" +import ( + "bytes" +) var ( // AVIF matches an AV1 Image File Format still or animated. // Wikipedia page seems outdated listing image/avif-sequence for animations. // https://github.com/AOMediaCodec/av1-avif/issues/59 AVIF = ftyp([]byte("avif"), []byte("avis")) - // Mp4 matches an MP4 file. - Mp4 = ftyp( - []byte("avc1"), []byte("dash"), []byte("iso2"), []byte("iso3"), - []byte("iso4"), []byte("iso5"), []byte("iso6"), []byte("isom"), - []byte("mmp4"), []byte("mp41"), []byte("mp42"), []byte("mp4v"), - []byte("mp71"), []byte("MSNV"), []byte("NDAS"), []byte("NDSC"), - []byte("NSDC"), []byte("NSDH"), []byte("NDSM"), []byte("NDSP"), - []byte("NDSS"), []byte("NDXC"), []byte("NDXH"), []byte("NDXM"), - []byte("NDXP"), []byte("NDXS"), []byte("F4V "), []byte("F4P "), - ) // ThreeGP matches a 3GPP file. ThreeGP = ftyp( []byte("3gp1"), []byte("3gp2"), []byte("3gp3"), []byte("3gp4"), @@ -53,6 +45,17 @@ var ( Heif = ftyp([]byte("mif1"), []byte("heim"), []byte("heis"), []byte("avic")) // HeifSequence matches a High Efficiency Image File Format (HEIF) file sequence. HeifSequence = ftyp([]byte("msf1"), []byte("hevm"), []byte("hevs"), []byte("avcs")) + // Mj2 matches a Motion JPEG 2000 file: https://en.wikipedia.org/wiki/Motion_JPEG_2000. + Mj2 = ftyp([]byte("mj2s"), []byte("mjp2"), []byte("MFSM"), []byte("MGSV")) + // Dvb matches a Digital Video Broadcasting file: https://dvb.org. + // https://cconcolato.github.io/mp4ra/filetype.html + // https://github.com/file/file/blob/512840337ead1076519332d24fefcaa8fac36e06/magic/Magdir/animation#L135-L154 + Dvb = ftyp( + []byte("dby1"), []byte("dsms"), []byte("dts1"), []byte("dts2"), + []byte("dts3"), []byte("dxo "), []byte("dmb1"), []byte("dmpf"), + []byte("drc1"), []byte("dv1a"), []byte("dv1b"), []byte("dv2a"), + []byte("dv2b"), []byte("dv3a"), []byte("dv3b"), []byte("dvr1"), + []byte("dvt1"), []byte("emsg")) // TODO: add support for remaining video formats at ftyps.com. ) @@ -86,3 +89,21 @@ func QuickTime(raw []byte, _ uint32) bool { } return bytes.Equal(raw[:8], []byte("\x00\x00\x00\x08wide")) } + +// Mp4 detects an .mp4 file. Mp4 detections only does a basic ftyp check. +// Mp4 has many registered and unregistered code points so it's hard to keep track +// of all. Detection will default on video/mp4 for all ftyp files. +// ISO_IEC_14496-12 is the specification for the iso container. +func Mp4(raw []byte, _ uint32) bool { + if len(raw) < 12 { + return false + } + // ftyps are made out of boxes. The first 4 bytes of the box represent + // its size in big-endian uint32. First box is the ftyp box and it is small + // in size. Check most significant byte is 0 to filter out false positive + // text files that happen to contain the string "ftyp" at index 4. + if raw[0] != 0 { + return false + } + return bytes.Equal(raw[4:8], []byte("ftyp")) +} diff --git a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/magic.go b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/magic.go index 3ce1de11..a34c6098 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/magic.go +++ b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/magic.go @@ -153,9 +153,6 @@ func ftyp(sigs ...[]byte) Detector { if len(raw) < 12 { return false } - if !bytes.Equal(raw[4:8], []byte("ftyp")) { - return false - } for _, s := range sigs { if bytes.Equal(raw[8:12], s) { return true @@ -242,3 +239,13 @@ func min(a, b int) int { } return b } + +type readBuf []byte + +func (b *readBuf) advance(n int) bool { + if n < 0 || len(*b) < n { + return false + } + *b = (*b)[n:] + return true +} diff --git a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ms_office.go b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ms_office.go index a1180173..7d60e22e 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ms_office.go +++ b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/ms_office.go @@ -5,58 +5,19 @@ import ( "encoding/binary" ) -var ( - xlsxSigFiles = [][]byte{ - []byte("xl/worksheets/"), - []byte("xl/drawings/"), - []byte("xl/theme/"), - []byte("xl/_rels/"), - []byte("xl/styles.xml"), - []byte("xl/workbook.xml"), - []byte("xl/sharedStrings.xml"), - } - docxSigFiles = [][]byte{ - []byte("word/media/"), - []byte("word/_rels/document.xml.rels"), - []byte("word/document.xml"), - []byte("word/styles.xml"), - []byte("word/fontTable.xml"), - []byte("word/settings.xml"), - []byte("word/numbering.xml"), - []byte("word/header"), - []byte("word/footer"), - } - pptxSigFiles = [][]byte{ - []byte("ppt/slides/"), - []byte("ppt/media/"), - []byte("ppt/slideLayouts/"), - []byte("ppt/theme/"), - []byte("ppt/slideMasters/"), - []byte("ppt/tags/"), - []byte("ppt/notesMasters/"), - []byte("ppt/_rels/"), - []byte("ppt/handoutMasters/"), - []byte("ppt/notesSlides/"), - []byte("ppt/presentation.xml"), - []byte("ppt/tableStyles.xml"), - []byte("ppt/presProps.xml"), - []byte("ppt/viewProps.xml"), - } -) - // Xlsx matches a Microsoft Excel 2007 file. func Xlsx(raw []byte, limit uint32) bool { - return zipContains(raw, xlsxSigFiles...) + return zipContains(raw, []byte("xl/"), true) } // Docx matches a Microsoft Word 2007 file. func Docx(raw []byte, limit uint32) bool { - return zipContains(raw, docxSigFiles...) + return zipContains(raw, []byte("word/"), true) } // Pptx matches a Microsoft PowerPoint 2007 file. func Pptx(raw []byte, limit uint32) bool { - return zipContains(raw, pptxSigFiles...) + return zipContains(raw, []byte("ppt/"), true) } // Ole matches an Open Linking and Embedding file. diff --git a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/text_csv.go b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/text_csv.go index af256438..6083ba8c 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/text_csv.go +++ b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/text_csv.go @@ -1,12 +1,28 @@ package magic import ( + "bufio" "bytes" "encoding/csv" "errors" "io" + "sync" ) +// A bufio.Reader pool to alleviate problems with memory allocations. +var readerPool = sync.Pool{ + New: func() any { + // Initiate with empty source reader. + return bufio.NewReader(nil) + }, +} + +func newReader(r io.Reader) *bufio.Reader { + br := readerPool.Get().(*bufio.Reader) + br.Reset(r) + return br +} + // Csv matches a comma-separated values file. func Csv(raw []byte, limit uint32) bool { return sv(raw, ',', limit) @@ -18,7 +34,11 @@ func Tsv(raw []byte, limit uint32) bool { } func sv(in []byte, comma rune, limit uint32) bool { - r := csv.NewReader(bytes.NewReader(dropLastLine(in, limit))) + in = dropLastLine(in, limit) + + br := newReader(bytes.NewReader(in)) + defer readerPool.Put(br) + r := csv.NewReader(br) r.Comma = comma r.ReuseRecord = true r.LazyQuotes = true diff --git a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/zip.go b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/zip.go index aaa27559..f866113f 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/internal/magic/zip.go +++ b/vendor/github.com/gabriel-vasile/mimetype/internal/magic/zip.go @@ -42,50 +42,71 @@ func Zip(raw []byte, limit uint32) bool { // Jar matches a Java archive file. func Jar(raw []byte, limit uint32) bool { - return zipContains(raw, []byte("META-INF/MANIFEST.MF")) + return zipContains(raw, []byte("META-INF/MANIFEST.MF"), false) } -// zipTokenizer holds the source zip file and scanned index. -type zipTokenizer struct { - in []byte - i int // current index -} +func zipContains(raw, sig []byte, msoCheck bool) bool { + b := readBuf(raw) + pk := []byte("PK\003\004") + if len(b) < 0x1E { + return false + } -// next returns the next file name from the zip headers. -// https://web.archive.org/web/20191129114319/https://users.cs.jmu.edu/buchhofp/forensics/formats/pkzip.html -func (t *zipTokenizer) next() (fileName []byte) { - if t.i > len(t.in) { - return + if !b.advance(0x1E) { + return false } - in := t.in[t.i:] - // pkSig is the signature of the zip local file header. - pkSig := []byte("PK\003\004") - pkIndex := bytes.Index(in, pkSig) - // 30 is the offset of the file name in the header. - fNameOffset := pkIndex + 30 - // end if signature not found or file name offset outside of file. - if pkIndex == -1 || fNameOffset > len(in) { - return + if bytes.HasPrefix(b, sig) { + return true } - fNameLen := int(binary.LittleEndian.Uint16(in[pkIndex+26 : pkIndex+28])) - if fNameLen <= 0 || fNameOffset+fNameLen > len(in) { - return - } - t.i += fNameOffset + fNameLen - return in[fNameOffset : fNameOffset+fNameLen] -} + if msoCheck { + skipFiles := [][]byte{ + []byte("[Content_Types].xml"), + []byte("_rels/.rels"), + []byte("docProps"), + []byte("customXml"), + []byte("[trash]"), + } -// zipContains returns true if the zip file headers from in contain any of the paths. -func zipContains(in []byte, paths ...[]byte) bool { - t := zipTokenizer{in: in} - for tok := t.next(); len(tok) != 0; tok = t.next() { - for p := range paths { - if bytes.HasPrefix(tok, paths[p]) { - return true + hasSkipFile := false + for _, sf := range skipFiles { + if bytes.HasPrefix(b, sf) { + hasSkipFile = true + break } } + if !hasSkipFile { + return false + } + } + + searchOffset := binary.LittleEndian.Uint32(raw[18:]) + 49 + if !b.advance(int(searchOffset)) { + return false } + nextHeader := bytes.Index(raw[searchOffset:], pk) + if !b.advance(nextHeader) { + return false + } + if bytes.HasPrefix(b, sig) { + return true + } + + for i := 0; i < 4; i++ { + if !b.advance(0x1A) { + return false + } + nextHeader = bytes.Index(b, pk) + if nextHeader == -1 { + return false + } + if !b.advance(nextHeader + 0x1E) { + return false + } + if bytes.HasPrefix(b, sig) { + return true + } + } return false } diff --git a/vendor/github.com/gabriel-vasile/mimetype/supported_mimes.md b/vendor/github.com/gabriel-vasile/mimetype/supported_mimes.md index 0a24ba38..042a5aa7 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/supported_mimes.md +++ b/vendor/github.com/gabriel-vasile/mimetype/supported_mimes.md @@ -1,4 +1,4 @@ -## 173 Supported MIME types +## 176 Supported MIME types This file is automatically generated when running tests. Do not edit manually. Extension | MIME type | Aliases @@ -75,21 +75,28 @@ Extension | MIME type | Aliases **.au** | audio/basic | - **.mpeg** | video/mpeg | - **.mov** | video/quicktime | - -**.mqv** | video/quicktime | - **.mp4** | video/mp4 | - -**.webm** | video/webm | audio/webm +**.avif** | image/avif | - **.3gp** | video/3gpp | video/3gp, audio/3gpp **.3g2** | video/3gpp2 | video/3g2, audio/3gpp2 +**.mp4** | audio/mp4 | audio/x-m4a, audio/x-mp4a +**.mqv** | video/quicktime | - +**.m4a** | audio/x-m4a | - +**.m4v** | video/x-m4v | - +**.heic** | image/heic | - +**.heic** | image/heic-sequence | - +**.heif** | image/heif | - +**.heif** | image/heif-sequence | - +**.mj2** | video/mj2 | - +**.dvb** | video/vnd.dvb.file | - +**.webm** | video/webm | audio/webm **.avi** | video/x-msvideo | video/avi, video/msvideo **.flv** | video/x-flv | - **.mkv** | video/x-matroska | - **.asf** | video/x-ms-asf | video/asf, video/x-ms-wmv **.aac** | audio/aac | - **.voc** | audio/x-unknown | - -**.mp4** | audio/mp4 | audio/x-m4a, audio/x-mp4a -**.m4a** | audio/x-m4a | - **.m3u** | application/vnd.apple.mpegurl | audio/mpegurl -**.m4v** | video/x-m4v | - **.rmvb** | application/vnd.rn-realmedia-vbr | - **.gz** | application/gzip | application/x-gzip, application/x-gunzip, application/gzipped, application/gzip-compressed, application/x-gzip-compressed, gzip/document **.class** | application/x-java-applet | - @@ -118,10 +125,6 @@ Extension | MIME type | Aliases **.macho** | application/x-mach-binary | - **.qcp** | audio/qcelp | - **.icns** | image/x-icns | - -**.heic** | image/heic | - -**.heic** | image/heic-sequence | - -**.heif** | image/heif | - -**.heif** | image/heif-sequence | - **.hdr** | image/vnd.radiance | - **.mrc** | application/marc | - **.mdb** | application/x-msaccess | - @@ -138,13 +141,13 @@ Extension | MIME type | Aliases **.pat** | image/x-gimp-pat | - **.gbr** | image/x-gimp-gbr | - **.glb** | model/gltf-binary | - -**.avif** | image/avif | - **.cab** | application/x-installshield | - **.jxr** | image/jxr | image/vnd.ms-photo +**.parquet** | application/vnd.apache.parquet | application/x-parquet **.txt** | text/plain | - **.html** | text/html | - **.svg** | image/svg+xml | - -**.xml** | text/xml | - +**.xml** | text/xml | application/xml **.rss** | application/rss+xml | text/rss **.atom** | application/atom+xml | - **.x3d** | model/x3d+xml | - diff --git a/vendor/github.com/gabriel-vasile/mimetype/tree.go b/vendor/github.com/gabriel-vasile/mimetype/tree.go index 75dcd7b1..aa7c8cdb 100644 --- a/vendor/github.com/gabriel-vasile/mimetype/tree.go +++ b/vendor/github.com/gabriel-vasile/mimetype/tree.go @@ -18,14 +18,13 @@ import ( var root = newMIME("application/octet-stream", "", func([]byte, uint32) bool { return true }, xpm, sevenZ, zip, pdf, fdf, ole, ps, psd, p7s, ogg, png, jpg, jxl, jp2, jpx, - jpm, jxs, gif, webp, exe, elf, ar, tar, xar, bz2, fits, tiff, bmp, ico, mp3, flac, - midi, ape, musePack, amr, wav, aiff, au, mpeg, quickTime, mqv, mp4, webM, - threeGP, threeG2, avi, flv, mkv, asf, aac, voc, aMp4, m4a, m3u, m4v, rmvb, - gzip, class, swf, crx, ttf, woff, woff2, otf, ttc, eot, wasm, shx, dbf, dcm, rar, - djvu, mobi, lit, bpg, sqlite3, dwg, nes, lnk, macho, qcp, icns, heic, - heicSeq, heif, heifSeq, hdr, mrc, mdb, accdb, zstd, cab, rpm, xz, lzip, - torrent, cpio, tzif, xcf, pat, gbr, glb, avif, cabIS, jxr, - // Keep text last because it is the slowest check + jpm, jxs, gif, webp, exe, elf, ar, tar, xar, bz2, fits, tiff, bmp, ico, mp3, + flac, midi, ape, musePack, amr, wav, aiff, au, mpeg, quickTime, mp4, webM, + avi, flv, mkv, asf, aac, voc, m3u, rmvb, gzip, class, swf, crx, ttf, woff, + woff2, otf, ttc, eot, wasm, shx, dbf, dcm, rar, djvu, mobi, lit, bpg, + sqlite3, dwg, nes, lnk, macho, qcp, icns, hdr, mrc, mdb, accdb, zstd, cab, + rpm, xz, lzip, torrent, cpio, tzif, xcf, pat, gbr, glb, cabIS, jxr, parquet, + // Keep text last because it is the slowest check. text, ) @@ -77,18 +76,19 @@ var ( oggAudio = newMIME("audio/ogg", ".oga", magic.OggAudio) oggVideo = newMIME("video/ogg", ".ogv", magic.OggVideo) text = newMIME("text/plain", ".txt", magic.Text, html, svg, xml, php, js, lua, perl, python, json, ndJSON, rtf, srt, tcl, csv, tsv, vCard, iCalendar, warc, vtt) - xml = newMIME("text/xml", ".xml", magic.XML, rss, atom, x3d, kml, xliff, collada, gml, gpx, tcx, amf, threemf, xfdf, owl2) - json = newMIME("application/json", ".json", magic.JSON, geoJSON, har) - har = newMIME("application/json", ".har", magic.HAR) - csv = newMIME("text/csv", ".csv", magic.Csv) - tsv = newMIME("text/tab-separated-values", ".tsv", magic.Tsv) - geoJSON = newMIME("application/geo+json", ".geojson", magic.GeoJSON) - ndJSON = newMIME("application/x-ndjson", ".ndjson", magic.NdJSON) - html = newMIME("text/html", ".html", magic.HTML) - php = newMIME("text/x-php", ".php", magic.Php) - rtf = newMIME("text/rtf", ".rtf", magic.Rtf).alias("application/rtf") - js = newMIME("application/javascript", ".js", magic.Js). - alias("application/x-javascript", "text/javascript") + xml = newMIME("text/xml", ".xml", magic.XML, rss, atom, x3d, kml, xliff, collada, gml, gpx, tcx, amf, threemf, xfdf, owl2). + alias("application/xml") + json = newMIME("application/json", ".json", magic.JSON, geoJSON, har) + har = newMIME("application/json", ".har", magic.HAR) + csv = newMIME("text/csv", ".csv", magic.Csv) + tsv = newMIME("text/tab-separated-values", ".tsv", magic.Tsv) + geoJSON = newMIME("application/geo+json", ".geojson", magic.GeoJSON) + ndJSON = newMIME("application/x-ndjson", ".ndjson", magic.NdJSON) + html = newMIME("text/html", ".html", magic.HTML) + php = newMIME("text/x-php", ".php", magic.Php) + rtf = newMIME("text/rtf", ".rtf", magic.Rtf).alias("application/rtf") + js = newMIME("application/javascript", ".js", magic.Js). + alias("application/x-javascript", "text/javascript") srt = newMIME("application/x-subrip", ".srt", magic.Srt). alias("application/x-srt", "text/x-srt") vtt = newMIME("text/vtt", ".vtt", magic.Vtt) @@ -161,7 +161,9 @@ var ( m3u = newMIME("application/vnd.apple.mpegurl", ".m3u", magic.M3u). alias("audio/mpegurl") m4v = newMIME("video/x-m4v", ".m4v", magic.M4v) - mp4 = newMIME("video/mp4", ".mp4", magic.Mp4) + mj2 = newMIME("video/mj2", ".mj2", magic.Mj2) + dvb = newMIME("video/vnd.dvb.file", ".dvb", magic.Dvb) + mp4 = newMIME("video/mp4", ".mp4", magic.Mp4, avif, threeGP, threeG2, aMp4, mqv, m4a, m4v, heic, heicSeq, heif, heifSeq, mj2, dvb) webM = newMIME("video/webm", ".webm", magic.WebM). alias("audio/webm") mpeg = newMIME("video/mpeg", ".mpeg", magic.Mpeg) @@ -257,4 +259,6 @@ var ( xfdf = newMIME("application/vnd.adobe.xfdf", ".xfdf", magic.Xfdf) glb = newMIME("model/gltf-binary", ".glb", magic.Glb) jxr = newMIME("image/jxr", ".jxr", magic.Jxr).alias("image/vnd.ms-photo") + parquet = newMIME("application/vnd.apache.parquet", ".parquet", magic.Par1). + alias("application/x-parquet") ) diff --git a/vendor/github.com/klauspost/compress/.goreleaser.yml b/vendor/github.com/klauspost/compress/.goreleaser.yml index a2295380..4528059c 100644 --- a/vendor/github.com/klauspost/compress/.goreleaser.yml +++ b/vendor/github.com/klauspost/compress/.goreleaser.yml @@ -1,5 +1,5 @@ -# This is an example goreleaser.yaml file with some sane defaults. -# Make sure to check the documentation at http://goreleaser.com +version: 2 + before: hooks: - ./gen.sh @@ -99,7 +99,7 @@ archives: checksum: name_template: 'checksums.txt' snapshot: - name_template: "{{ .Tag }}-next" + version_template: "{{ .Tag }}-next" changelog: sort: asc filters: diff --git a/vendor/github.com/klauspost/compress/README.md b/vendor/github.com/klauspost/compress/README.md index 684a3085..de264c85 100644 --- a/vendor/github.com/klauspost/compress/README.md +++ b/vendor/github.com/klauspost/compress/README.md @@ -16,6 +16,13 @@ This package provides various compression algorithms. # changelog +* Sep 23rd, 2024 - [1.17.10](https://github.com/klauspost/compress/releases/tag/v1.17.10) + * gzhttp: Add TransportAlwaysDecompress option. https://github.com/klauspost/compress/pull/978 + * gzhttp: Add supported decompress request body by @mirecl in https://github.com/klauspost/compress/pull/1002 + * s2: Add EncodeBuffer buffer recycling callback https://github.com/klauspost/compress/pull/982 + * zstd: Improve memory usage on small streaming encodes https://github.com/klauspost/compress/pull/1007 + * flate: read data written with partial flush by @vajexal in https://github.com/klauspost/compress/pull/996 + * Jun 12th, 2024 - [1.17.9](https://github.com/klauspost/compress/releases/tag/v1.17.9) * s2: Reduce ReadFrom temporary allocations https://github.com/klauspost/compress/pull/949 * flate, zstd: Shave some bytes off amd64 matchLen by @greatroar in https://github.com/klauspost/compress/pull/963 diff --git a/vendor/github.com/klauspost/compress/s2/encode.go b/vendor/github.com/klauspost/compress/s2/encode.go index 0c9088ad..20b80227 100644 --- a/vendor/github.com/klauspost/compress/s2/encode.go +++ b/vendor/github.com/klauspost/compress/s2/encode.go @@ -9,6 +9,9 @@ import ( "encoding/binary" "math" "math/bits" + "sync" + + "github.com/klauspost/compress/internal/race" ) // Encode returns the encoded form of src. The returned slice may be a sub- @@ -52,6 +55,8 @@ func Encode(dst, src []byte) []byte { return dst[:d] } +var estblockPool [2]sync.Pool + // EstimateBlockSize will perform a very fast compression // without outputting the result and return the compressed output size. // The function returns -1 if no improvement could be achieved. @@ -61,9 +66,25 @@ func EstimateBlockSize(src []byte) (d int) { return -1 } if len(src) <= 1024 { - d = calcBlockSizeSmall(src) + const sz, pool = 2048, 0 + tmp, ok := estblockPool[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer estblockPool[pool].Put(tmp) + + d = calcBlockSizeSmall(src, tmp) } else { - d = calcBlockSize(src) + const sz, pool = 32768, 1 + tmp, ok := estblockPool[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer estblockPool[pool].Put(tmp) + + d = calcBlockSize(src, tmp) } if d == 0 { diff --git a/vendor/github.com/klauspost/compress/s2/encode_amd64.go b/vendor/github.com/klauspost/compress/s2/encode_amd64.go index 4f45206a..7aadd255 100644 --- a/vendor/github.com/klauspost/compress/s2/encode_amd64.go +++ b/vendor/github.com/klauspost/compress/s2/encode_amd64.go @@ -3,10 +3,16 @@ package s2 -import "github.com/klauspost/compress/internal/race" +import ( + "sync" + + "github.com/klauspost/compress/internal/race" +) const hasAmd64Asm = true +var encPools [4]sync.Pool + // encodeBlock encodes a non-empty src to a guaranteed-large-enough dst. It // assumes that the varint-encoded length of the decompressed bytes has already // been written. @@ -29,23 +35,60 @@ func encodeBlock(dst, src []byte) (d int) { ) if len(src) >= 4<<20 { - return encodeBlockAsm(dst, src) + const sz, pool = 65536, 0 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeBlockAsm(dst, src, tmp) } if len(src) >= limit12B { - return encodeBlockAsm4MB(dst, src) + const sz, pool = 65536, 0 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeBlockAsm4MB(dst, src, tmp) } if len(src) >= limit10B { - return encodeBlockAsm12B(dst, src) + const sz, pool = 16384, 1 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeBlockAsm12B(dst, src, tmp) } if len(src) >= limit8B { - return encodeBlockAsm10B(dst, src) + const sz, pool = 4096, 2 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeBlockAsm10B(dst, src, tmp) } if len(src) < minNonLiteralBlockSize { return 0 } - return encodeBlockAsm8B(dst, src) + const sz, pool = 1024, 3 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeBlockAsm8B(dst, src, tmp) } +var encBetterPools [5]sync.Pool + // encodeBlockBetter encodes a non-empty src to a guaranteed-large-enough dst. It // assumes that the varint-encoded length of the decompressed bytes has already // been written. @@ -68,21 +111,59 @@ func encodeBlockBetter(dst, src []byte) (d int) { ) if len(src) > 4<<20 { - return encodeBetterBlockAsm(dst, src) + const sz, pool = 589824, 0 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + return encodeBetterBlockAsm(dst, src, tmp) } if len(src) >= limit12B { - return encodeBetterBlockAsm4MB(dst, src) + const sz, pool = 589824, 0 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + + return encodeBetterBlockAsm4MB(dst, src, tmp) } if len(src) >= limit10B { - return encodeBetterBlockAsm12B(dst, src) + const sz, pool = 81920, 0 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + + return encodeBetterBlockAsm12B(dst, src, tmp) } if len(src) >= limit8B { - return encodeBetterBlockAsm10B(dst, src) + const sz, pool = 20480, 1 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + return encodeBetterBlockAsm10B(dst, src, tmp) } if len(src) < minNonLiteralBlockSize { return 0 } - return encodeBetterBlockAsm8B(dst, src) + + const sz, pool = 5120, 2 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + return encodeBetterBlockAsm8B(dst, src, tmp) } // encodeBlockSnappy encodes a non-empty src to a guaranteed-large-enough dst. It @@ -105,22 +186,57 @@ func encodeBlockSnappy(dst, src []byte) (d int) { // Use 8 bit table when less than... limit8B = 512 ) - if len(src) >= 64<<10 { - return encodeSnappyBlockAsm(dst, src) + if len(src) > 65536 { + const sz, pool = 65536, 0 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeSnappyBlockAsm(dst, src, tmp) } if len(src) >= limit12B { - return encodeSnappyBlockAsm64K(dst, src) + const sz, pool = 65536, 0 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeSnappyBlockAsm64K(dst, src, tmp) } if len(src) >= limit10B { - return encodeSnappyBlockAsm12B(dst, src) + const sz, pool = 16384, 1 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeSnappyBlockAsm12B(dst, src, tmp) } if len(src) >= limit8B { - return encodeSnappyBlockAsm10B(dst, src) + const sz, pool = 4096, 2 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeSnappyBlockAsm10B(dst, src, tmp) } if len(src) < minNonLiteralBlockSize { return 0 } - return encodeSnappyBlockAsm8B(dst, src) + const sz, pool = 1024, 3 + tmp, ok := encPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encPools[pool].Put(tmp) + return encodeSnappyBlockAsm8B(dst, src, tmp) } // encodeBlockSnappy encodes a non-empty src to a guaranteed-large-enough dst. It @@ -143,20 +259,59 @@ func encodeBlockBetterSnappy(dst, src []byte) (d int) { // Use 8 bit table when less than... limit8B = 512 ) - if len(src) >= 64<<10 { - return encodeSnappyBetterBlockAsm(dst, src) + if len(src) > 65536 { + const sz, pool = 589824, 0 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + return encodeSnappyBetterBlockAsm(dst, src, tmp) } + if len(src) >= limit12B { - return encodeSnappyBetterBlockAsm64K(dst, src) + const sz, pool = 294912, 4 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + + return encodeSnappyBetterBlockAsm64K(dst, src, tmp) } if len(src) >= limit10B { - return encodeSnappyBetterBlockAsm12B(dst, src) + const sz, pool = 81920, 0 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + + return encodeSnappyBetterBlockAsm12B(dst, src, tmp) } if len(src) >= limit8B { - return encodeSnappyBetterBlockAsm10B(dst, src) + const sz, pool = 20480, 1 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + return encodeSnappyBetterBlockAsm10B(dst, src, tmp) } if len(src) < minNonLiteralBlockSize { return 0 } - return encodeSnappyBetterBlockAsm8B(dst, src) + + const sz, pool = 5120, 2 + tmp, ok := encBetterPools[pool].Get().(*[sz]byte) + if !ok { + tmp = &[sz]byte{} + } + race.WriteSlice(tmp[:]) + defer encBetterPools[pool].Put(tmp) + return encodeSnappyBetterBlockAsm8B(dst, src, tmp) } diff --git a/vendor/github.com/klauspost/compress/s2/encode_go.go b/vendor/github.com/klauspost/compress/s2/encode_go.go index 6b393c34..dd1c973c 100644 --- a/vendor/github.com/klauspost/compress/s2/encode_go.go +++ b/vendor/github.com/klauspost/compress/s2/encode_go.go @@ -317,7 +317,7 @@ func matchLen(a []byte, b []byte) int { } // input must be > inputMargin -func calcBlockSize(src []byte) (d int) { +func calcBlockSize(src []byte, _ *[32768]byte) (d int) { // Initialize the hash table. const ( tableBits = 13 @@ -503,7 +503,7 @@ emitRemainder: } // length must be > inputMargin. -func calcBlockSizeSmall(src []byte) (d int) { +func calcBlockSizeSmall(src []byte, _ *[2048]byte) (d int) { // Initialize the hash table. const ( tableBits = 9 diff --git a/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.go b/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.go index 297e4150..f43aa815 100644 --- a/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.go +++ b/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.go @@ -11,154 +11,154 @@ func _dummy_() // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBlockAsm(dst []byte, src []byte) int +func encodeBlockAsm(dst []byte, src []byte, tmp *[65536]byte) int // encodeBlockAsm4MB encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4194304 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBlockAsm4MB(dst []byte, src []byte) int +func encodeBlockAsm4MB(dst []byte, src []byte, tmp *[65536]byte) int // encodeBlockAsm12B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 16383 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBlockAsm12B(dst []byte, src []byte) int +func encodeBlockAsm12B(dst []byte, src []byte, tmp *[16384]byte) int // encodeBlockAsm10B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4095 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBlockAsm10B(dst []byte, src []byte) int +func encodeBlockAsm10B(dst []byte, src []byte, tmp *[4096]byte) int // encodeBlockAsm8B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 511 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBlockAsm8B(dst []byte, src []byte) int +func encodeBlockAsm8B(dst []byte, src []byte, tmp *[1024]byte) int // encodeBetterBlockAsm encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4294967295 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBetterBlockAsm(dst []byte, src []byte) int +func encodeBetterBlockAsm(dst []byte, src []byte, tmp *[589824]byte) int // encodeBetterBlockAsm4MB encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4194304 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBetterBlockAsm4MB(dst []byte, src []byte) int +func encodeBetterBlockAsm4MB(dst []byte, src []byte, tmp *[589824]byte) int // encodeBetterBlockAsm12B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 16383 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBetterBlockAsm12B(dst []byte, src []byte) int +func encodeBetterBlockAsm12B(dst []byte, src []byte, tmp *[81920]byte) int // encodeBetterBlockAsm10B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4095 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBetterBlockAsm10B(dst []byte, src []byte) int +func encodeBetterBlockAsm10B(dst []byte, src []byte, tmp *[20480]byte) int // encodeBetterBlockAsm8B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 511 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeBetterBlockAsm8B(dst []byte, src []byte) int +func encodeBetterBlockAsm8B(dst []byte, src []byte, tmp *[5120]byte) int // encodeSnappyBlockAsm encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4294967295 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBlockAsm(dst []byte, src []byte) int +func encodeSnappyBlockAsm(dst []byte, src []byte, tmp *[65536]byte) int // encodeSnappyBlockAsm64K encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 65535 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBlockAsm64K(dst []byte, src []byte) int +func encodeSnappyBlockAsm64K(dst []byte, src []byte, tmp *[65536]byte) int // encodeSnappyBlockAsm12B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 16383 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBlockAsm12B(dst []byte, src []byte) int +func encodeSnappyBlockAsm12B(dst []byte, src []byte, tmp *[16384]byte) int // encodeSnappyBlockAsm10B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4095 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBlockAsm10B(dst []byte, src []byte) int +func encodeSnappyBlockAsm10B(dst []byte, src []byte, tmp *[4096]byte) int // encodeSnappyBlockAsm8B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 511 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBlockAsm8B(dst []byte, src []byte) int +func encodeSnappyBlockAsm8B(dst []byte, src []byte, tmp *[1024]byte) int // encodeSnappyBetterBlockAsm encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4294967295 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBetterBlockAsm(dst []byte, src []byte) int +func encodeSnappyBetterBlockAsm(dst []byte, src []byte, tmp *[589824]byte) int // encodeSnappyBetterBlockAsm64K encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 65535 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBetterBlockAsm64K(dst []byte, src []byte) int +func encodeSnappyBetterBlockAsm64K(dst []byte, src []byte, tmp *[294912]byte) int // encodeSnappyBetterBlockAsm12B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 16383 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBetterBlockAsm12B(dst []byte, src []byte) int +func encodeSnappyBetterBlockAsm12B(dst []byte, src []byte, tmp *[81920]byte) int // encodeSnappyBetterBlockAsm10B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4095 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBetterBlockAsm10B(dst []byte, src []byte) int +func encodeSnappyBetterBlockAsm10B(dst []byte, src []byte, tmp *[20480]byte) int // encodeSnappyBetterBlockAsm8B encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 511 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func encodeSnappyBetterBlockAsm8B(dst []byte, src []byte) int +func encodeSnappyBetterBlockAsm8B(dst []byte, src []byte, tmp *[5120]byte) int // calcBlockSize encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 4294967295 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func calcBlockSize(src []byte) int +func calcBlockSize(src []byte, tmp *[32768]byte) int // calcBlockSizeSmall encodes a non-empty src to a guaranteed-large-enough dst. // Maximum input 1024 bytes. // It assumes that the varint-encoded length of the decompressed bytes has already been written. // //go:noescape -func calcBlockSizeSmall(src []byte) int +func calcBlockSizeSmall(src []byte, tmp *[2048]byte) int // emitLiteral writes a literal chunk and returns the number of bytes written. // diff --git a/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.s b/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.s index 2ff5b334..df9be687 100644 --- a/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.s +++ b/vendor/github.com/klauspost/compress/s2/encodeblock_amd64.s @@ -13,1270 +13,1271 @@ TEXT ·_dummy_(SB), $0 #endif RET -// func encodeBlockAsm(dst []byte, src []byte) int +// func encodeBlockAsm(dst []byte, src []byte, tmp *[65536]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBlockAsm(SB), $65560-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000200, CX - LEAQ 24(SP), DX +TEXT ·encodeBlockAsm(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBlockAsm: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBlockAsm MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeBlockAsm: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x06, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x06, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBlockAsm - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x10, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x10, R11 + IMULQ R9, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeBlockAsm - LEAL 1(CX), SI - MOVL 12(SP), DI - MOVL SI, BX - SUBL 16(SP), BX + LEAL 1(DX), DI + MOVL 12(SP), R8 + MOVL DI, SI + SUBL 16(SP), SI JZ repeat_extend_back_end_encodeBlockAsm repeat_extend_back_loop_encodeBlockAsm: - CMPL SI, DI + CMPL DI, R8 JBE repeat_extend_back_end_encodeBlockAsm - MOVB -1(DX)(BX*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(SI*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeBlockAsm - LEAL -1(SI), SI - DECL BX + LEAL -1(DI), DI + DECL SI JNZ repeat_extend_back_loop_encodeBlockAsm repeat_extend_back_end_encodeBlockAsm: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 5(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 5(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeBlockAsm: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeBlockAsm - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeBlockAsm - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeBlockAsm - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_repeat_emit_encodeBlockAsm - CMPL BX, $0x01000000 + CMPL SI, $0x01000000 JB four_bytes_repeat_emit_encodeBlockAsm - MOVB $0xfc, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP memmove_long_repeat_emit_encodeBlockAsm four_bytes_repeat_emit_encodeBlockAsm: - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_repeat_emit_encodeBlockAsm three_bytes_repeat_emit_encodeBlockAsm: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeBlockAsm two_bytes_repeat_emit_encodeBlockAsm: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeBlockAsm JMP memmove_long_repeat_emit_encodeBlockAsm one_byte_repeat_emit_encodeBlockAsm: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_repeat_emit_encodeBlockAsm emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm emit_lit_memmove_repeat_emit_encodeBlockAsm_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_repeat_emit_encodeBlockAsm: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeBlockAsm memmove_long_repeat_emit_encodeBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R11 - SHRQ $0x05, R11 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R12 - SUBQ R10, R12 - DECQ R11 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R12 + SHRQ $0x05, R12 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R13 + SUBQ R11, R13 + DECQ R12 JA emit_lit_memmove_long_repeat_emit_encodeBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(R9)(R12*1), R10 - LEAQ -32(AX)(R12*1), R13 + LEAQ -32(R10)(R13*1), R11 + LEAQ -32(CX)(R13*1), R14 emit_lit_memmove_long_repeat_emit_encodeBlockAsmlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R13) - MOVOA X5, 16(R13) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R14) + MOVOA X5, 16(R14) + ADDQ $0x20, R14 + ADDQ $0x20, R11 ADDQ $0x20, R13 - ADDQ $0x20, R10 - ADDQ $0x20, R12 - DECQ R11 + DECQ R12 JNA emit_lit_memmove_long_repeat_emit_encodeBlockAsmlarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(R9)(R12*1), X4 - MOVOU -16(R9)(R12*1), X5 - MOVOA X4, -32(AX)(R12*1) - MOVOA X5, -16(AX)(R12*1) - ADDQ $0x20, R12 - CMPQ R8, R12 + MOVOU -32(R10)(R13*1), X4 + MOVOU -16(R10)(R13*1), X5 + MOVOA X4, -32(CX)(R13*1) + MOVOA X5, -16(CX)(R13*1) + ADDQ $0x20, R13 + CMPQ R9, R13 JAE emit_lit_memmove_long_repeat_emit_encodeBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeBlockAsm: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), R8 - SUBL CX, R8 - LEAQ (DX)(CX*1), R9 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R9 + SUBL DX, R9 + LEAQ (BX)(DX*1), R10 + LEAQ (BX)(SI*1), SI // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_repeat_extend_encodeBlockAsm: - CMPL R8, $0x10 + CMPL R9, $0x10 JB matchlen_match8_repeat_extend_encodeBlockAsm - MOVQ (R9)(R11*1), R10 - MOVQ 8(R9)(R11*1), R12 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + MOVQ 8(R10)(R12*1), R13 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm - XORQ 8(BX)(R11*1), R12 + XORQ 8(SI)(R12*1), R13 JNZ matchlen_bsf_16repeat_extend_encodeBlockAsm - LEAL -16(R8), R8 - LEAL 16(R11), R11 + LEAL -16(R9), R9 + LEAL 16(R12), R12 JMP matchlen_loopback_16_repeat_extend_encodeBlockAsm matchlen_bsf_16repeat_extend_encodeBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm matchlen_match8_repeat_extend_encodeBlockAsm: - CMPL R8, $0x08 + CMPL R9, $0x08 JB matchlen_match4_repeat_extend_encodeBlockAsm - MOVQ (R9)(R11*1), R10 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm - LEAL -8(R8), R8 - LEAL 8(R11), R11 + LEAL -8(R9), R9 + LEAL 8(R12), R12 JMP matchlen_match4_repeat_extend_encodeBlockAsm matchlen_bsf_8_repeat_extend_encodeBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm matchlen_match4_repeat_extend_encodeBlockAsm: - CMPL R8, $0x04 + CMPL R9, $0x04 JB matchlen_match2_repeat_extend_encodeBlockAsm - MOVL (R9)(R11*1), R10 - CMPL (BX)(R11*1), R10 + MOVL (R10)(R12*1), R11 + CMPL (SI)(R12*1), R11 JNE matchlen_match2_repeat_extend_encodeBlockAsm - LEAL -4(R8), R8 - LEAL 4(R11), R11 + LEAL -4(R9), R9 + LEAL 4(R12), R12 matchlen_match2_repeat_extend_encodeBlockAsm: - CMPL R8, $0x01 + CMPL R9, $0x01 JE matchlen_match1_repeat_extend_encodeBlockAsm JB repeat_extend_forward_end_encodeBlockAsm - MOVW (R9)(R11*1), R10 - CMPW (BX)(R11*1), R10 + MOVW (R10)(R12*1), R11 + CMPW (SI)(R12*1), R11 JNE matchlen_match1_repeat_extend_encodeBlockAsm - LEAL 2(R11), R11 - SUBL $0x02, R8 + LEAL 2(R12), R12 + SUBL $0x02, R9 JZ repeat_extend_forward_end_encodeBlockAsm matchlen_match1_repeat_extend_encodeBlockAsm: - MOVB (R9)(R11*1), R10 - CMPB (BX)(R11*1), R10 + MOVB (R10)(R12*1), R11 + CMPB (SI)(R12*1), R11 JNE repeat_extend_forward_end_encodeBlockAsm - LEAL 1(R11), R11 + LEAL 1(R12), R12 repeat_extend_forward_end_encodeBlockAsm: - ADDL R11, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI - TESTL DI, DI + ADDL R12, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI + TESTL R8, R8 JZ repeat_as_copy_encodeBlockAsm // emitRepeat emit_repeat_again_match_repeat_encodeBlockAsm: - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_match_repeat_encodeBlockAsm - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_match_repeat_encodeBlockAsm - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_match_repeat_encodeBlockAsm cant_repeat_two_offset_match_repeat_encodeBlockAsm: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_match_repeat_encodeBlockAsm - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_match_repeat_encodeBlockAsm - CMPL BX, $0x0100ffff + CMPL SI, $0x0100ffff JB repeat_five_match_repeat_encodeBlockAsm - LEAL -16842747(BX), BX - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(SI), SI + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_repeat_encodeBlockAsm repeat_five_match_repeat_encodeBlockAsm: - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm repeat_four_match_repeat_encodeBlockAsm: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm repeat_three_match_repeat_encodeBlockAsm: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_match_repeat_encodeBlockAsm: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_offset_match_repeat_encodeBlockAsm: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm repeat_as_copy_encodeBlockAsm: // emitCopy - CMPL SI, $0x00010000 + CMPL DI, $0x00010000 JB two_byte_offset_repeat_as_copy_encodeBlockAsm - CMPL BX, $0x40 + CMPL SI, $0x40 JBE four_bytes_remain_repeat_as_copy_encodeBlockAsm - MOVB $0xff, (AX) - MOVL SI, 1(AX) - LEAL -64(BX), BX - ADDQ $0x05, AX - CMPL BX, $0x04 + MOVB $0xff, (CX) + MOVL DI, 1(CX) + LEAL -64(SI), SI + ADDQ $0x05, CX + CMPL SI, $0x04 JB four_bytes_remain_repeat_as_copy_encodeBlockAsm // emitRepeat emit_repeat_again_repeat_as_copy_encodeBlockAsm_emit_copy: - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm_emit_copy - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm_emit_copy - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_repeat_as_copy_encodeBlockAsm_emit_copy - CMPL BX, $0x0100ffff + CMPL SI, $0x0100ffff JB repeat_five_repeat_as_copy_encodeBlockAsm_emit_copy - LEAL -16842747(BX), BX - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(SI), SI + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_repeat_as_copy_encodeBlockAsm_emit_copy repeat_five_repeat_as_copy_encodeBlockAsm_emit_copy: - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm repeat_four_repeat_as_copy_encodeBlockAsm_emit_copy: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm repeat_three_repeat_as_copy_encodeBlockAsm_emit_copy: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_repeat_as_copy_encodeBlockAsm_emit_copy: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm four_bytes_remain_repeat_as_copy_encodeBlockAsm: - TESTL BX, BX + TESTL SI, SI JZ repeat_end_emit_encodeBlockAsm - XORL DI, DI - LEAL -1(DI)(BX*4), BX - MOVB BL, (AX) - MOVL SI, 1(AX) - ADDQ $0x05, AX + XORL R8, R8 + LEAL -1(R8)(SI*4), SI + MOVB SI, (CX) + MOVL DI, 1(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm two_byte_offset_repeat_as_copy_encodeBlockAsm: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeBlockAsm - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE long_offset_short_repeat_as_copy_encodeBlockAsm - MOVL $0x00000001, DI - LEAL 16(DI), DI - MOVB SI, 1(AX) - MOVL SI, R8 - SHRL $0x08, R8 - SHLL $0x05, R8 - ORL R8, DI - MOVB DI, (AX) - ADDQ $0x02, AX - SUBL $0x08, BX + MOVL $0x00000001, R8 + LEAL 16(R8), R8 + MOVB DI, 1(CX) + MOVL DI, R9 + SHRL $0x08, R9 + SHLL $0x05, R9 + ORL R9, R8 + MOVB R8, (CX) + ADDQ $0x02, CX + SUBL $0x08, SI // emitRepeat - LEAL -4(BX), BX + LEAL -4(SI), SI JMP cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b emit_repeat_again_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b - CMPL BX, $0x0100ffff + CMPL SI, $0x0100ffff JB repeat_five_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b - LEAL -16842747(BX), BX - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(SI), SI + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b repeat_five_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm repeat_four_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm repeat_three_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short_2b: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm long_offset_short_repeat_as_copy_encodeBlockAsm: - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX // emitRepeat emit_repeat_again_repeat_as_copy_encodeBlockAsm_emit_copy_short: - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm_emit_copy_short - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm_emit_copy_short - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_repeat_as_copy_encodeBlockAsm_emit_copy_short - CMPL BX, $0x0100ffff + CMPL SI, $0x0100ffff JB repeat_five_repeat_as_copy_encodeBlockAsm_emit_copy_short - LEAL -16842747(BX), BX - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(SI), SI + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_repeat_as_copy_encodeBlockAsm_emit_copy_short repeat_five_repeat_as_copy_encodeBlockAsm_emit_copy_short: - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm repeat_four_repeat_as_copy_encodeBlockAsm_emit_copy_short: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm repeat_three_repeat_as_copy_encodeBlockAsm_emit_copy_short: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_repeat_as_copy_encodeBlockAsm_emit_copy_short: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm repeat_two_offset_repeat_as_copy_encodeBlockAsm_emit_copy_short: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm two_byte_offset_short_repeat_as_copy_encodeBlockAsm: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeBlockAsm - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeBlockAsm - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm emit_copy_three_repeat_as_copy_encodeBlockAsm: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeBlockAsm: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeBlockAsm no_repeat_found_encodeBlockAsm: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBlockAsm - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeBlockAsm - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeBlockAsm - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBlockAsm candidate3_match_encodeBlockAsm: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeBlockAsm candidate2_match_encodeBlockAsm: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeBlockAsm: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBlockAsm match_extend_back_loop_encodeBlockAsm: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBlockAsm - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBlockAsm - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBlockAsm JMP match_extend_back_loop_encodeBlockAsm match_extend_back_end_encodeBlockAsm: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 5(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 5(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBlockAsm: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeBlockAsm - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeBlockAsm - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeBlockAsm - CMPL DI, $0x00010000 + CMPL R8, $0x00010000 JB three_bytes_match_emit_encodeBlockAsm - CMPL DI, $0x01000000 + CMPL R8, $0x01000000 JB four_bytes_match_emit_encodeBlockAsm - MOVB $0xfc, (AX) - MOVL DI, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL R8, 1(CX) + ADDQ $0x05, CX JMP memmove_long_match_emit_encodeBlockAsm four_bytes_match_emit_encodeBlockAsm: - MOVL DI, R9 - SHRL $0x10, R9 - MOVB $0xf8, (AX) - MOVW DI, 1(AX) - MOVB R9, 3(AX) - ADDQ $0x04, AX + MOVL R8, R10 + SHRL $0x10, R10 + MOVB $0xf8, (CX) + MOVW R8, 1(CX) + MOVB R10, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_encodeBlockAsm three_bytes_match_emit_encodeBlockAsm: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBlockAsm two_bytes_match_emit_encodeBlockAsm: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeBlockAsm JMP memmove_long_match_emit_encodeBlockAsm one_byte_match_emit_encodeBlockAsm: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBlockAsm: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeBlockAsm emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm emit_lit_memmove_match_emit_encodeBlockAsm_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBlockAsm: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeBlockAsm memmove_long_match_emit_encodeBlockAsm: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeBlockAsmlarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeBlockAsmlarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeBlockAsm: match_nolit_loop_encodeBlockAsm: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeBlockAsm: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeBlockAsm - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeBlockAsm - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeBlockAsm matchlen_bsf_16match_nolit_encodeBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeBlockAsm matchlen_match8_match_nolit_encodeBlockAsm: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeBlockAsm - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeBlockAsm matchlen_bsf_8_match_nolit_encodeBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeBlockAsm matchlen_match4_match_nolit_encodeBlockAsm: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeBlockAsm - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeBlockAsm - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeBlockAsm: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeBlockAsm JB match_nolit_end_encodeBlockAsm - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeBlockAsm - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeBlockAsm matchlen_match1_match_nolit_encodeBlockAsm: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeBlockAsm - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeBlockAsm: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB two_byte_offset_match_nolit_encodeBlockAsm - CMPL R9, $0x40 + CMPL R10, $0x40 JBE four_bytes_remain_match_nolit_encodeBlockAsm - MOVB $0xff, (AX) - MOVL BX, 1(AX) - LEAL -64(R9), R9 - ADDQ $0x05, AX - CMPL R9, $0x04 + MOVB $0xff, (CX) + MOVL SI, 1(CX) + LEAL -64(R10), R10 + ADDQ $0x05, CX + CMPL R10, $0x04 JB four_bytes_remain_match_nolit_encodeBlockAsm // emitRepeat emit_repeat_again_match_nolit_encodeBlockAsm_emit_copy: - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm_emit_copy - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm_emit_copy - CMPL R9, $0x00010100 + CMPL R10, $0x00010100 JB repeat_four_match_nolit_encodeBlockAsm_emit_copy - CMPL R9, $0x0100ffff + CMPL R10, $0x0100ffff JB repeat_five_match_nolit_encodeBlockAsm_emit_copy - LEAL -16842747(R9), R9 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R10), R10 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_encodeBlockAsm_emit_copy repeat_five_match_nolit_encodeBlockAsm_emit_copy: - LEAL -65536(R9), R9 - MOVL R9, BX - MOVW $0x001d, (AX) - MOVW R9, 2(AX) - SARL $0x10, BX - MOVB BL, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R10), R10 + MOVL R10, SI + MOVW $0x001d, (CX) + MOVW R10, 2(CX) + SARL $0x10, SI + MOVB SI, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_four_match_nolit_encodeBlockAsm_emit_copy: - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_three_match_nolit_encodeBlockAsm_emit_copy: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_two_match_nolit_encodeBlockAsm_emit_copy: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm four_bytes_remain_match_nolit_encodeBlockAsm: - TESTL R9, R9 + TESTL R10, R10 JZ match_nolit_emitcopy_end_encodeBlockAsm - XORL SI, SI - LEAL -1(SI)(R9*4), R9 - MOVB R9, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + XORL DI, DI + LEAL -1(DI)(R10*4), R10 + MOVB R10, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm two_byte_offset_match_nolit_encodeBlockAsm: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeBlockAsm - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE long_offset_short_match_nolit_encodeBlockAsm - MOVL $0x00000001, SI - LEAL 16(SI), SI - MOVB BL, 1(AX) - MOVL BX, DI - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, SI - MOVB SI, (AX) - ADDQ $0x02, AX - SUBL $0x08, R9 + MOVL $0x00000001, DI + LEAL 16(DI), DI + MOVB SI, 1(CX) + MOVL SI, R8 + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, DI + MOVB DI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R10 // emitRepeat - LEAL -4(R9), R9 + LEAL -4(R10), R10 JMP cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short_2b emit_repeat_again_match_nolit_encodeBlockAsm_emit_copy_short_2b: - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm_emit_copy_short_2b - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short_2b - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short_2b: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm_emit_copy_short_2b - CMPL R9, $0x00010100 + CMPL R10, $0x00010100 JB repeat_four_match_nolit_encodeBlockAsm_emit_copy_short_2b - CMPL R9, $0x0100ffff + CMPL R10, $0x0100ffff JB repeat_five_match_nolit_encodeBlockAsm_emit_copy_short_2b - LEAL -16842747(R9), R9 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R10), R10 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_encodeBlockAsm_emit_copy_short_2b repeat_five_match_nolit_encodeBlockAsm_emit_copy_short_2b: - LEAL -65536(R9), R9 - MOVL R9, BX - MOVW $0x001d, (AX) - MOVW R9, 2(AX) - SARL $0x10, BX - MOVB BL, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R10), R10 + MOVL R10, SI + MOVW $0x001d, (CX) + MOVW R10, 2(CX) + SARL $0x10, SI + MOVB SI, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_four_match_nolit_encodeBlockAsm_emit_copy_short_2b: - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_three_match_nolit_encodeBlockAsm_emit_copy_short_2b: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_two_match_nolit_encodeBlockAsm_emit_copy_short_2b: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short_2b: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm long_offset_short_match_nolit_encodeBlockAsm: - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX // emitRepeat emit_repeat_again_match_nolit_encodeBlockAsm_emit_copy_short: - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm_emit_copy_short - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm_emit_copy_short - CMPL R9, $0x00010100 + CMPL R10, $0x00010100 JB repeat_four_match_nolit_encodeBlockAsm_emit_copy_short - CMPL R9, $0x0100ffff + CMPL R10, $0x0100ffff JB repeat_five_match_nolit_encodeBlockAsm_emit_copy_short - LEAL -16842747(R9), R9 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R10), R10 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_encodeBlockAsm_emit_copy_short repeat_five_match_nolit_encodeBlockAsm_emit_copy_short: - LEAL -65536(R9), R9 - MOVL R9, BX - MOVW $0x001d, (AX) - MOVW R9, 2(AX) - SARL $0x10, BX - MOVB BL, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R10), R10 + MOVL R10, SI + MOVW $0x001d, (CX) + MOVW R10, 2(CX) + SARL $0x10, SI + MOVB SI, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_four_match_nolit_encodeBlockAsm_emit_copy_short: - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_three_match_nolit_encodeBlockAsm_emit_copy_short: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_two_match_nolit_encodeBlockAsm_emit_copy_short: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm repeat_two_offset_match_nolit_encodeBlockAsm_emit_copy_short: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm two_byte_offset_short_match_nolit_encodeBlockAsm: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeBlockAsm - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeBlockAsm - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm emit_copy_three_match_nolit_encodeBlockAsm: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeBlockAsm: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBlockAsm - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBlockAsm: - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x10, DI - IMULQ R8, DI - SHRQ $0x32, DI - SHLQ $0x10, BX - IMULQ R8, BX - SHRQ $0x32, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x10, R8 + IMULQ R9, R8 + SHRQ $0x32, R8 + SHLQ $0x10, SI + IMULQ R9, SI + SHRQ $0x32, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeBlockAsm - INCL CX + INCL DX JMP search_loop_encodeBlockAsm emit_remainder_encodeBlockAsm: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 5(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 5(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBlockAsm: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBlockAsm - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBlockAsm @@ -1286,41 +1287,41 @@ emit_remainder_ok_encodeBlockAsm: JB three_bytes_emit_remainder_encodeBlockAsm CMPL DX, $0x01000000 JB four_bytes_emit_remainder_encodeBlockAsm - MOVB $0xfc, (AX) - MOVL DX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL DX, 1(CX) + ADDQ $0x05, CX JMP memmove_long_emit_remainder_encodeBlockAsm four_bytes_emit_remainder_encodeBlockAsm: MOVL DX, BX SHRL $0x10, BX - MOVB $0xf8, (AX) - MOVW DX, 1(AX) - MOVB BL, 3(AX) - ADDQ $0x04, AX + MOVB $0xf8, (CX) + MOVW DX, 1(CX) + MOVB BL, 3(CX) + ADDQ $0x04, CX JMP memmove_long_emit_remainder_encodeBlockAsm three_bytes_emit_remainder_encodeBlockAsm: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBlockAsm two_bytes_emit_remainder_encodeBlockAsm: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBlockAsm JMP memmove_long_emit_remainder_encodeBlockAsm one_byte_emit_remainder_encodeBlockAsm: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -1336,73 +1337,73 @@ memmove_emit_remainder_encodeBlockAsm: JMP emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBlockAsm emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm emit_lit_memmove_emit_remainder_encodeBlockAsm_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBlockAsm: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBlockAsm memmove_long_emit_remainder_encodeBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBlockAsmlarge_big_loop_back: MOVOU (SI), X4 @@ -1416,1199 +1417,1200 @@ emit_lit_memmove_long_emit_remainder_encodeBlockAsmlarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBlockAsmlarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBlockAsm: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBlockAsm4MB(dst []byte, src []byte) int +// func encodeBlockAsm4MB(dst []byte, src []byte, tmp *[65536]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBlockAsm4MB(SB), $65560-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000200, CX - LEAQ 24(SP), DX +TEXT ·encodeBlockAsm4MB(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBlockAsm4MB: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBlockAsm4MB MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeBlockAsm4MB: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x06, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x06, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBlockAsm4MB - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x10, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x10, R11 + IMULQ R9, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeBlockAsm4MB - LEAL 1(CX), SI - MOVL 12(SP), DI - MOVL SI, BX - SUBL 16(SP), BX + LEAL 1(DX), DI + MOVL 12(SP), R8 + MOVL DI, SI + SUBL 16(SP), SI JZ repeat_extend_back_end_encodeBlockAsm4MB repeat_extend_back_loop_encodeBlockAsm4MB: - CMPL SI, DI + CMPL DI, R8 JBE repeat_extend_back_end_encodeBlockAsm4MB - MOVB -1(DX)(BX*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(SI*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeBlockAsm4MB - LEAL -1(SI), SI - DECL BX + LEAL -1(DI), DI + DECL SI JNZ repeat_extend_back_loop_encodeBlockAsm4MB repeat_extend_back_end_encodeBlockAsm4MB: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 4(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 4(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeBlockAsm4MB: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeBlockAsm4MB - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeBlockAsm4MB - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeBlockAsm4MB - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_repeat_emit_encodeBlockAsm4MB - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_repeat_emit_encodeBlockAsm4MB three_bytes_repeat_emit_encodeBlockAsm4MB: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeBlockAsm4MB two_bytes_repeat_emit_encodeBlockAsm4MB: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeBlockAsm4MB JMP memmove_long_repeat_emit_encodeBlockAsm4MB one_byte_repeat_emit_encodeBlockAsm4MB: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeBlockAsm4MB: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_repeat_emit_encodeBlockAsm4MB emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm4MB emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm4MB emit_lit_memmove_repeat_emit_encodeBlockAsm4MB_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_repeat_emit_encodeBlockAsm4MB: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeBlockAsm4MB memmove_long_repeat_emit_encodeBlockAsm4MB: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R11 - SHRQ $0x05, R11 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R12 - SUBQ R10, R12 - DECQ R11 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R12 + SHRQ $0x05, R12 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R13 + SUBQ R11, R13 + DECQ R12 JA emit_lit_memmove_long_repeat_emit_encodeBlockAsm4MBlarge_forward_sse_loop_32 - LEAQ -32(R9)(R12*1), R10 - LEAQ -32(AX)(R12*1), R13 + LEAQ -32(R10)(R13*1), R11 + LEAQ -32(CX)(R13*1), R14 emit_lit_memmove_long_repeat_emit_encodeBlockAsm4MBlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R13) - MOVOA X5, 16(R13) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R14) + MOVOA X5, 16(R14) + ADDQ $0x20, R14 + ADDQ $0x20, R11 ADDQ $0x20, R13 - ADDQ $0x20, R10 - ADDQ $0x20, R12 - DECQ R11 + DECQ R12 JNA emit_lit_memmove_long_repeat_emit_encodeBlockAsm4MBlarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeBlockAsm4MBlarge_forward_sse_loop_32: - MOVOU -32(R9)(R12*1), X4 - MOVOU -16(R9)(R12*1), X5 - MOVOA X4, -32(AX)(R12*1) - MOVOA X5, -16(AX)(R12*1) - ADDQ $0x20, R12 - CMPQ R8, R12 + MOVOU -32(R10)(R13*1), X4 + MOVOU -16(R10)(R13*1), X5 + MOVOA X4, -32(CX)(R13*1) + MOVOA X5, -16(CX)(R13*1) + ADDQ $0x20, R13 + CMPQ R9, R13 JAE emit_lit_memmove_long_repeat_emit_encodeBlockAsm4MBlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeBlockAsm4MB: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), R8 - SUBL CX, R8 - LEAQ (DX)(CX*1), R9 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R9 + SUBL DX, R9 + LEAQ (BX)(DX*1), R10 + LEAQ (BX)(SI*1), SI // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_repeat_extend_encodeBlockAsm4MB: - CMPL R8, $0x10 + CMPL R9, $0x10 JB matchlen_match8_repeat_extend_encodeBlockAsm4MB - MOVQ (R9)(R11*1), R10 - MOVQ 8(R9)(R11*1), R12 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + MOVQ 8(R10)(R12*1), R13 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm4MB - XORQ 8(BX)(R11*1), R12 + XORQ 8(SI)(R12*1), R13 JNZ matchlen_bsf_16repeat_extend_encodeBlockAsm4MB - LEAL -16(R8), R8 - LEAL 16(R11), R11 + LEAL -16(R9), R9 + LEAL 16(R12), R12 JMP matchlen_loopback_16_repeat_extend_encodeBlockAsm4MB matchlen_bsf_16repeat_extend_encodeBlockAsm4MB: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm4MB matchlen_match8_repeat_extend_encodeBlockAsm4MB: - CMPL R8, $0x08 + CMPL R9, $0x08 JB matchlen_match4_repeat_extend_encodeBlockAsm4MB - MOVQ (R9)(R11*1), R10 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm4MB - LEAL -8(R8), R8 - LEAL 8(R11), R11 + LEAL -8(R9), R9 + LEAL 8(R12), R12 JMP matchlen_match4_repeat_extend_encodeBlockAsm4MB matchlen_bsf_8_repeat_extend_encodeBlockAsm4MB: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm4MB matchlen_match4_repeat_extend_encodeBlockAsm4MB: - CMPL R8, $0x04 + CMPL R9, $0x04 JB matchlen_match2_repeat_extend_encodeBlockAsm4MB - MOVL (R9)(R11*1), R10 - CMPL (BX)(R11*1), R10 + MOVL (R10)(R12*1), R11 + CMPL (SI)(R12*1), R11 JNE matchlen_match2_repeat_extend_encodeBlockAsm4MB - LEAL -4(R8), R8 - LEAL 4(R11), R11 + LEAL -4(R9), R9 + LEAL 4(R12), R12 matchlen_match2_repeat_extend_encodeBlockAsm4MB: - CMPL R8, $0x01 + CMPL R9, $0x01 JE matchlen_match1_repeat_extend_encodeBlockAsm4MB JB repeat_extend_forward_end_encodeBlockAsm4MB - MOVW (R9)(R11*1), R10 - CMPW (BX)(R11*1), R10 + MOVW (R10)(R12*1), R11 + CMPW (SI)(R12*1), R11 JNE matchlen_match1_repeat_extend_encodeBlockAsm4MB - LEAL 2(R11), R11 - SUBL $0x02, R8 + LEAL 2(R12), R12 + SUBL $0x02, R9 JZ repeat_extend_forward_end_encodeBlockAsm4MB matchlen_match1_repeat_extend_encodeBlockAsm4MB: - MOVB (R9)(R11*1), R10 - CMPB (BX)(R11*1), R10 + MOVB (R10)(R12*1), R11 + CMPB (SI)(R12*1), R11 JNE repeat_extend_forward_end_encodeBlockAsm4MB - LEAL 1(R11), R11 + LEAL 1(R12), R12 repeat_extend_forward_end_encodeBlockAsm4MB: - ADDL R11, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI - TESTL DI, DI + ADDL R12, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI + TESTL R8, R8 JZ repeat_as_copy_encodeBlockAsm4MB // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_match_repeat_encodeBlockAsm4MB - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_match_repeat_encodeBlockAsm4MB - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_match_repeat_encodeBlockAsm4MB cant_repeat_two_offset_match_repeat_encodeBlockAsm4MB: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_match_repeat_encodeBlockAsm4MB - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_match_repeat_encodeBlockAsm4MB - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_four_match_repeat_encodeBlockAsm4MB: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_three_match_repeat_encodeBlockAsm4MB: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_match_repeat_encodeBlockAsm4MB: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_offset_match_repeat_encodeBlockAsm4MB: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_as_copy_encodeBlockAsm4MB: // emitCopy - CMPL SI, $0x00010000 + CMPL DI, $0x00010000 JB two_byte_offset_repeat_as_copy_encodeBlockAsm4MB - CMPL BX, $0x40 + CMPL SI, $0x40 JBE four_bytes_remain_repeat_as_copy_encodeBlockAsm4MB - MOVB $0xff, (AX) - MOVL SI, 1(AX) - LEAL -64(BX), BX - ADDQ $0x05, AX - CMPL BX, $0x04 + MOVB $0xff, (CX) + MOVL DI, 1(CX) + LEAL -64(SI), SI + ADDQ $0x05, CX + CMPL SI, $0x04 JB four_bytes_remain_repeat_as_copy_encodeBlockAsm4MB // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm4MB_emit_copy - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm4MB_emit_copy - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_repeat_as_copy_encodeBlockAsm4MB_emit_copy - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_four_repeat_as_copy_encodeBlockAsm4MB_emit_copy: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_three_repeat_as_copy_encodeBlockAsm4MB_emit_copy: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_repeat_as_copy_encodeBlockAsm4MB_emit_copy: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB four_bytes_remain_repeat_as_copy_encodeBlockAsm4MB: - TESTL BX, BX + TESTL SI, SI JZ repeat_end_emit_encodeBlockAsm4MB - XORL DI, DI - LEAL -1(DI)(BX*4), BX - MOVB BL, (AX) - MOVL SI, 1(AX) - ADDQ $0x05, AX + XORL R8, R8 + LEAL -1(R8)(SI*4), SI + MOVB SI, (CX) + MOVL DI, 1(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm4MB two_byte_offset_repeat_as_copy_encodeBlockAsm4MB: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeBlockAsm4MB - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE long_offset_short_repeat_as_copy_encodeBlockAsm4MB - MOVL $0x00000001, DI - LEAL 16(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX - SUBL $0x08, BX + MOVL $0x00000001, R8 + LEAL 16(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX + SUBL $0x08, SI // emitRepeat - LEAL -4(BX), BX + LEAL -4(SI), SI JMP cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_four_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_three_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short_2b: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB long_offset_short_repeat_as_copy_encodeBlockAsm4MB: - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short - CMPL BX, $0x00010100 + CMPL SI, $0x00010100 JB repeat_four_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short - LEAL -65536(BX), BX - MOVL BX, SI - MOVW $0x001d, (AX) - MOVW BX, 2(AX) - SARL $0x10, SI - MOVB SI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(SI), SI + MOVL SI, DI + MOVW $0x001d, (CX) + MOVW SI, 2(CX) + SARL $0x10, DI + MOVB DI, 4(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_four_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short: - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_three_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB repeat_two_offset_repeat_as_copy_encodeBlockAsm4MB_emit_copy_short: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB two_byte_offset_short_repeat_as_copy_encodeBlockAsm4MB: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeBlockAsm4MB - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeBlockAsm4MB - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm4MB emit_copy_three_repeat_as_copy_encodeBlockAsm4MB: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeBlockAsm4MB: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeBlockAsm4MB no_repeat_found_encodeBlockAsm4MB: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBlockAsm4MB - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeBlockAsm4MB - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeBlockAsm4MB - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBlockAsm4MB candidate3_match_encodeBlockAsm4MB: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeBlockAsm4MB candidate2_match_encodeBlockAsm4MB: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeBlockAsm4MB: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBlockAsm4MB match_extend_back_loop_encodeBlockAsm4MB: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBlockAsm4MB - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBlockAsm4MB - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBlockAsm4MB JMP match_extend_back_loop_encodeBlockAsm4MB match_extend_back_end_encodeBlockAsm4MB: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 4(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 4(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBlockAsm4MB: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeBlockAsm4MB - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeBlockAsm4MB - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeBlockAsm4MB - CMPL DI, $0x00010000 + CMPL R8, $0x00010000 JB three_bytes_match_emit_encodeBlockAsm4MB - MOVL DI, R9 - SHRL $0x10, R9 - MOVB $0xf8, (AX) - MOVW DI, 1(AX) - MOVB R9, 3(AX) - ADDQ $0x04, AX + MOVL R8, R10 + SHRL $0x10, R10 + MOVB $0xf8, (CX) + MOVW R8, 1(CX) + MOVB R10, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_encodeBlockAsm4MB three_bytes_match_emit_encodeBlockAsm4MB: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBlockAsm4MB two_bytes_match_emit_encodeBlockAsm4MB: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeBlockAsm4MB JMP memmove_long_match_emit_encodeBlockAsm4MB one_byte_match_emit_encodeBlockAsm4MB: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBlockAsm4MB: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeBlockAsm4MB emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm4MB emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm4MB emit_lit_memmove_match_emit_encodeBlockAsm4MB_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBlockAsm4MB: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeBlockAsm4MB memmove_long_match_emit_encodeBlockAsm4MB: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeBlockAsm4MBlarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeBlockAsm4MBlarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeBlockAsm4MBlarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBlockAsm4MBlarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeBlockAsm4MBlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeBlockAsm4MB: match_nolit_loop_encodeBlockAsm4MB: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeBlockAsm4MB: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeBlockAsm4MB - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm4MB - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeBlockAsm4MB - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeBlockAsm4MB matchlen_bsf_16match_nolit_encodeBlockAsm4MB: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeBlockAsm4MB matchlen_match8_match_nolit_encodeBlockAsm4MB: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeBlockAsm4MB - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm4MB - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeBlockAsm4MB matchlen_bsf_8_match_nolit_encodeBlockAsm4MB: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeBlockAsm4MB matchlen_match4_match_nolit_encodeBlockAsm4MB: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeBlockAsm4MB - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeBlockAsm4MB - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeBlockAsm4MB: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeBlockAsm4MB JB match_nolit_end_encodeBlockAsm4MB - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeBlockAsm4MB - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeBlockAsm4MB matchlen_match1_match_nolit_encodeBlockAsm4MB: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeBlockAsm4MB - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeBlockAsm4MB: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB two_byte_offset_match_nolit_encodeBlockAsm4MB - CMPL R9, $0x40 + CMPL R10, $0x40 JBE four_bytes_remain_match_nolit_encodeBlockAsm4MB - MOVB $0xff, (AX) - MOVL BX, 1(AX) - LEAL -64(R9), R9 - ADDQ $0x05, AX - CMPL R9, $0x04 + MOVB $0xff, (CX) + MOVL SI, 1(CX) + LEAL -64(R10), R10 + ADDQ $0x05, CX + CMPL R10, $0x04 JB four_bytes_remain_match_nolit_encodeBlockAsm4MB // emitRepeat - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm4MB_emit_copy - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm4MB_emit_copy - CMPL R9, $0x00010100 + CMPL R10, $0x00010100 JB repeat_four_match_nolit_encodeBlockAsm4MB_emit_copy - LEAL -65536(R9), R9 - MOVL R9, BX - MOVW $0x001d, (AX) - MOVW R9, 2(AX) - SARL $0x10, BX - MOVB BL, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R10), R10 + MOVL R10, SI + MOVW $0x001d, (CX) + MOVW R10, 2(CX) + SARL $0x10, SI + MOVB SI, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_four_match_nolit_encodeBlockAsm4MB_emit_copy: - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_three_match_nolit_encodeBlockAsm4MB_emit_copy: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_two_match_nolit_encodeBlockAsm4MB_emit_copy: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB four_bytes_remain_match_nolit_encodeBlockAsm4MB: - TESTL R9, R9 + TESTL R10, R10 JZ match_nolit_emitcopy_end_encodeBlockAsm4MB - XORL SI, SI - LEAL -1(SI)(R9*4), R9 - MOVB R9, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + XORL DI, DI + LEAL -1(DI)(R10*4), R10 + MOVB R10, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB two_byte_offset_match_nolit_encodeBlockAsm4MB: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeBlockAsm4MB - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE long_offset_short_match_nolit_encodeBlockAsm4MB - MOVL $0x00000001, SI - LEAL 16(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX - SUBL $0x08, R9 - + MOVL $0x00000001, DI + LEAL 16(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R10 + // emitRepeat - LEAL -4(R9), R9 + LEAL -4(R10), R10 JMP cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b - CMPL R9, $0x00010100 + CMPL R10, $0x00010100 JB repeat_four_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b - LEAL -65536(R9), R9 - MOVL R9, BX - MOVW $0x001d, (AX) - MOVW R9, 2(AX) - SARL $0x10, BX - MOVB BL, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R10), R10 + MOVL R10, SI + MOVW $0x001d, (CX) + MOVW R10, 2(CX) + SARL $0x10, SI + MOVB SI, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_four_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b: - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_three_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_two_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short_2b: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB long_offset_short_match_nolit_encodeBlockAsm4MB: - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX // emitRepeat - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm4MB_emit_copy_short - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm4MB_emit_copy_short - CMPL R9, $0x00010100 + CMPL R10, $0x00010100 JB repeat_four_match_nolit_encodeBlockAsm4MB_emit_copy_short - LEAL -65536(R9), R9 - MOVL R9, BX - MOVW $0x001d, (AX) - MOVW R9, 2(AX) - SARL $0x10, BX - MOVB BL, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R10), R10 + MOVL R10, SI + MOVW $0x001d, (CX) + MOVW R10, 2(CX) + SARL $0x10, SI + MOVB SI, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_four_match_nolit_encodeBlockAsm4MB_emit_copy_short: - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_three_match_nolit_encodeBlockAsm4MB_emit_copy_short: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_two_match_nolit_encodeBlockAsm4MB_emit_copy_short: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB repeat_two_offset_match_nolit_encodeBlockAsm4MB_emit_copy_short: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB two_byte_offset_short_match_nolit_encodeBlockAsm4MB: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeBlockAsm4MB - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeBlockAsm4MB - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm4MB emit_copy_three_match_nolit_encodeBlockAsm4MB: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeBlockAsm4MB: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBlockAsm4MB - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBlockAsm4MB: - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x10, DI - IMULQ R8, DI - SHRQ $0x32, DI - SHLQ $0x10, BX - IMULQ R8, BX - SHRQ $0x32, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x10, R8 + IMULQ R9, R8 + SHRQ $0x32, R8 + SHLQ $0x10, SI + IMULQ R9, SI + SHRQ $0x32, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeBlockAsm4MB - INCL CX + INCL DX JMP search_loop_encodeBlockAsm4MB emit_remainder_encodeBlockAsm4MB: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 4(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 4(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBlockAsm4MB: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBlockAsm4MB - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBlockAsm4MB @@ -2618,33 +2620,33 @@ emit_remainder_ok_encodeBlockAsm4MB: JB three_bytes_emit_remainder_encodeBlockAsm4MB MOVL DX, BX SHRL $0x10, BX - MOVB $0xf8, (AX) - MOVW DX, 1(AX) - MOVB BL, 3(AX) - ADDQ $0x04, AX + MOVB $0xf8, (CX) + MOVW DX, 1(CX) + MOVB BL, 3(CX) + ADDQ $0x04, CX JMP memmove_long_emit_remainder_encodeBlockAsm4MB three_bytes_emit_remainder_encodeBlockAsm4MB: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBlockAsm4MB two_bytes_emit_remainder_encodeBlockAsm4MB: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBlockAsm4MB JMP memmove_long_emit_remainder_encodeBlockAsm4MB one_byte_emit_remainder_encodeBlockAsm4MB: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBlockAsm4MB: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -2660,73 +2662,73 @@ memmove_emit_remainder_encodeBlockAsm4MB: JMP emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBlockAsm4MB_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBlockAsm4MB: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBlockAsm4MB memmove_long_emit_remainder_encodeBlockAsm4MB: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBlockAsm4MBlarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBlockAsm4MBlarge_big_loop_back: MOVOU (SI), X4 @@ -2740,967 +2742,968 @@ emit_lit_memmove_long_emit_remainder_encodeBlockAsm4MBlarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBlockAsm4MBlarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBlockAsm4MBlarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBlockAsm4MBlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBlockAsm4MB: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBlockAsm12B(dst []byte, src []byte) int +// func encodeBlockAsm12B(dst []byte, src []byte, tmp *[16384]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBlockAsm12B(SB), $16408-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000080, CX - LEAQ 24(SP), DX +TEXT ·encodeBlockAsm12B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000080, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBlockAsm12B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBlockAsm12B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeBlockAsm12B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBlockAsm12B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x000000cf1bbcdcbb, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x18, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x000000cf1bbcdcbb, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x18, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x34, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x18, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x18, R11 + IMULQ R9, R11 + SHRQ $0x34, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x18, R10 + IMULQ R9, R10 + SHRQ $0x34, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeBlockAsm12B - LEAL 1(CX), SI - MOVL 12(SP), DI - MOVL SI, BX - SUBL 16(SP), BX + LEAL 1(DX), DI + MOVL 12(SP), R8 + MOVL DI, SI + SUBL 16(SP), SI JZ repeat_extend_back_end_encodeBlockAsm12B repeat_extend_back_loop_encodeBlockAsm12B: - CMPL SI, DI + CMPL DI, R8 JBE repeat_extend_back_end_encodeBlockAsm12B - MOVB -1(DX)(BX*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(SI*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeBlockAsm12B - LEAL -1(SI), SI - DECL BX + LEAL -1(DI), DI + DECL SI JNZ repeat_extend_back_loop_encodeBlockAsm12B repeat_extend_back_end_encodeBlockAsm12B: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeBlockAsm12B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeBlockAsm12B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeBlockAsm12B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeBlockAsm12B JB three_bytes_repeat_emit_encodeBlockAsm12B three_bytes_repeat_emit_encodeBlockAsm12B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeBlockAsm12B two_bytes_repeat_emit_encodeBlockAsm12B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeBlockAsm12B JMP memmove_long_repeat_emit_encodeBlockAsm12B one_byte_repeat_emit_encodeBlockAsm12B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_repeat_emit_encodeBlockAsm12B emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm12B emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm12B emit_lit_memmove_repeat_emit_encodeBlockAsm12B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_repeat_emit_encodeBlockAsm12B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeBlockAsm12B memmove_long_repeat_emit_encodeBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R11 - SHRQ $0x05, R11 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R12 - SUBQ R10, R12 - DECQ R11 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R12 + SHRQ $0x05, R12 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R13 + SUBQ R11, R13 + DECQ R12 JA emit_lit_memmove_long_repeat_emit_encodeBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R12*1), R10 - LEAQ -32(AX)(R12*1), R13 + LEAQ -32(R10)(R13*1), R11 + LEAQ -32(CX)(R13*1), R14 emit_lit_memmove_long_repeat_emit_encodeBlockAsm12Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R13) - MOVOA X5, 16(R13) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R14) + MOVOA X5, 16(R14) + ADDQ $0x20, R14 + ADDQ $0x20, R11 ADDQ $0x20, R13 - ADDQ $0x20, R10 - ADDQ $0x20, R12 - DECQ R11 + DECQ R12 JNA emit_lit_memmove_long_repeat_emit_encodeBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R12*1), X4 - MOVOU -16(R9)(R12*1), X5 - MOVOA X4, -32(AX)(R12*1) - MOVOA X5, -16(AX)(R12*1) - ADDQ $0x20, R12 - CMPQ R8, R12 + MOVOU -32(R10)(R13*1), X4 + MOVOU -16(R10)(R13*1), X5 + MOVOA X4, -32(CX)(R13*1) + MOVOA X5, -16(CX)(R13*1) + ADDQ $0x20, R13 + CMPQ R9, R13 JAE emit_lit_memmove_long_repeat_emit_encodeBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeBlockAsm12B: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), R8 - SUBL CX, R8 - LEAQ (DX)(CX*1), R9 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R9 + SUBL DX, R9 + LEAQ (BX)(DX*1), R10 + LEAQ (BX)(SI*1), SI // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_repeat_extend_encodeBlockAsm12B: - CMPL R8, $0x10 + CMPL R9, $0x10 JB matchlen_match8_repeat_extend_encodeBlockAsm12B - MOVQ (R9)(R11*1), R10 - MOVQ 8(R9)(R11*1), R12 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + MOVQ 8(R10)(R12*1), R13 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm12B - XORQ 8(BX)(R11*1), R12 + XORQ 8(SI)(R12*1), R13 JNZ matchlen_bsf_16repeat_extend_encodeBlockAsm12B - LEAL -16(R8), R8 - LEAL 16(R11), R11 + LEAL -16(R9), R9 + LEAL 16(R12), R12 JMP matchlen_loopback_16_repeat_extend_encodeBlockAsm12B matchlen_bsf_16repeat_extend_encodeBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm12B matchlen_match8_repeat_extend_encodeBlockAsm12B: - CMPL R8, $0x08 + CMPL R9, $0x08 JB matchlen_match4_repeat_extend_encodeBlockAsm12B - MOVQ (R9)(R11*1), R10 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm12B - LEAL -8(R8), R8 - LEAL 8(R11), R11 + LEAL -8(R9), R9 + LEAL 8(R12), R12 JMP matchlen_match4_repeat_extend_encodeBlockAsm12B matchlen_bsf_8_repeat_extend_encodeBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm12B matchlen_match4_repeat_extend_encodeBlockAsm12B: - CMPL R8, $0x04 + CMPL R9, $0x04 JB matchlen_match2_repeat_extend_encodeBlockAsm12B - MOVL (R9)(R11*1), R10 - CMPL (BX)(R11*1), R10 + MOVL (R10)(R12*1), R11 + CMPL (SI)(R12*1), R11 JNE matchlen_match2_repeat_extend_encodeBlockAsm12B - LEAL -4(R8), R8 - LEAL 4(R11), R11 + LEAL -4(R9), R9 + LEAL 4(R12), R12 matchlen_match2_repeat_extend_encodeBlockAsm12B: - CMPL R8, $0x01 + CMPL R9, $0x01 JE matchlen_match1_repeat_extend_encodeBlockAsm12B JB repeat_extend_forward_end_encodeBlockAsm12B - MOVW (R9)(R11*1), R10 - CMPW (BX)(R11*1), R10 + MOVW (R10)(R12*1), R11 + CMPW (SI)(R12*1), R11 JNE matchlen_match1_repeat_extend_encodeBlockAsm12B - LEAL 2(R11), R11 - SUBL $0x02, R8 + LEAL 2(R12), R12 + SUBL $0x02, R9 JZ repeat_extend_forward_end_encodeBlockAsm12B matchlen_match1_repeat_extend_encodeBlockAsm12B: - MOVB (R9)(R11*1), R10 - CMPB (BX)(R11*1), R10 + MOVB (R10)(R12*1), R11 + CMPB (SI)(R12*1), R11 JNE repeat_extend_forward_end_encodeBlockAsm12B - LEAL 1(R11), R11 + LEAL 1(R12), R12 repeat_extend_forward_end_encodeBlockAsm12B: - ADDL R11, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI - TESTL DI, DI + ADDL R12, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI + TESTL R8, R8 JZ repeat_as_copy_encodeBlockAsm12B // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_match_repeat_encodeBlockAsm12B - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_match_repeat_encodeBlockAsm12B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_match_repeat_encodeBlockAsm12B cant_repeat_two_offset_match_repeat_encodeBlockAsm12B: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_match_repeat_encodeBlockAsm12B - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_three_match_repeat_encodeBlockAsm12B: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_two_match_repeat_encodeBlockAsm12B: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_two_offset_match_repeat_encodeBlockAsm12B: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_as_copy_encodeBlockAsm12B: // emitCopy - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeBlockAsm12B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE long_offset_short_repeat_as_copy_encodeBlockAsm12B - MOVL $0x00000001, DI - LEAL 16(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX - SUBL $0x08, BX + MOVL $0x00000001, R8 + LEAL 16(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX + SUBL $0x08, SI // emitRepeat - LEAL -4(BX), BX + LEAL -4(SI), SI JMP cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_three_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_two_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short_2b: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B long_offset_short_repeat_as_copy_encodeBlockAsm12B: - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm12B_emit_copy_short - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm12B_emit_copy_short - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_three_repeat_as_copy_encodeBlockAsm12B_emit_copy_short: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_two_repeat_as_copy_encodeBlockAsm12B_emit_copy_short: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B repeat_two_offset_repeat_as_copy_encodeBlockAsm12B_emit_copy_short: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B two_byte_offset_short_repeat_as_copy_encodeBlockAsm12B: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeBlockAsm12B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeBlockAsm12B - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm12B emit_copy_three_repeat_as_copy_encodeBlockAsm12B: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeBlockAsm12B: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeBlockAsm12B no_repeat_found_encodeBlockAsm12B: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBlockAsm12B - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeBlockAsm12B - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeBlockAsm12B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBlockAsm12B candidate3_match_encodeBlockAsm12B: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeBlockAsm12B candidate2_match_encodeBlockAsm12B: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeBlockAsm12B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBlockAsm12B match_extend_back_loop_encodeBlockAsm12B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBlockAsm12B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBlockAsm12B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBlockAsm12B JMP match_extend_back_loop_encodeBlockAsm12B match_extend_back_end_encodeBlockAsm12B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBlockAsm12B: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeBlockAsm12B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeBlockAsm12B - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeBlockAsm12B JB three_bytes_match_emit_encodeBlockAsm12B three_bytes_match_emit_encodeBlockAsm12B: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBlockAsm12B two_bytes_match_emit_encodeBlockAsm12B: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeBlockAsm12B JMP memmove_long_match_emit_encodeBlockAsm12B one_byte_match_emit_encodeBlockAsm12B: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBlockAsm12B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeBlockAsm12B emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm12B emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm12B emit_lit_memmove_match_emit_encodeBlockAsm12B_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBlockAsm12B: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeBlockAsm12B memmove_long_match_emit_encodeBlockAsm12B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeBlockAsm12Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeBlockAsm12B: match_nolit_loop_encodeBlockAsm12B: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeBlockAsm12B: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeBlockAsm12B - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm12B - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeBlockAsm12B - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeBlockAsm12B matchlen_bsf_16match_nolit_encodeBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeBlockAsm12B matchlen_match8_match_nolit_encodeBlockAsm12B: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeBlockAsm12B - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm12B - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeBlockAsm12B matchlen_bsf_8_match_nolit_encodeBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeBlockAsm12B matchlen_match4_match_nolit_encodeBlockAsm12B: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeBlockAsm12B - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeBlockAsm12B - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeBlockAsm12B: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeBlockAsm12B JB match_nolit_end_encodeBlockAsm12B - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeBlockAsm12B - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeBlockAsm12B matchlen_match1_match_nolit_encodeBlockAsm12B: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeBlockAsm12B - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeBlockAsm12B: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeBlockAsm12B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE long_offset_short_match_nolit_encodeBlockAsm12B - MOVL $0x00000001, SI - LEAL 16(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX - SUBL $0x08, R9 + MOVL $0x00000001, DI + LEAL 16(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R10 // emitRepeat - LEAL -4(R9), R9 + LEAL -4(R10), R10 JMP cant_repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short_2b - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm12B_emit_copy_short_2b - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short_2b - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short_2b: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm12B_emit_copy_short_2b - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B repeat_three_match_nolit_encodeBlockAsm12B_emit_copy_short_2b: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B repeat_two_match_nolit_encodeBlockAsm12B_emit_copy_short_2b: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short_2b: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B long_offset_short_match_nolit_encodeBlockAsm12B: - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX // emitRepeat - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm12B_emit_copy_short - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm12B_emit_copy_short - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B repeat_three_match_nolit_encodeBlockAsm12B_emit_copy_short: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B repeat_two_match_nolit_encodeBlockAsm12B_emit_copy_short: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B repeat_two_offset_match_nolit_encodeBlockAsm12B_emit_copy_short: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX - JMP match_nolit_emitcopy_end_encodeBlockAsm12B - + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX + JMP match_nolit_emitcopy_end_encodeBlockAsm12B + two_byte_offset_short_match_nolit_encodeBlockAsm12B: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeBlockAsm12B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeBlockAsm12B - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm12B emit_copy_three_match_nolit_encodeBlockAsm12B: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeBlockAsm12B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBlockAsm12B - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBlockAsm12B: - MOVQ $0x000000cf1bbcdcbb, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x18, DI - IMULQ R8, DI - SHRQ $0x34, DI - SHLQ $0x18, BX - IMULQ R8, BX - SHRQ $0x34, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x000000cf1bbcdcbb, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x18, R8 + IMULQ R9, R8 + SHRQ $0x34, R8 + SHLQ $0x18, SI + IMULQ R9, SI + SHRQ $0x34, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeBlockAsm12B - INCL CX + INCL DX JMP search_loop_encodeBlockAsm12B emit_remainder_encodeBlockAsm12B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBlockAsm12B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBlockAsm12B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBlockAsm12B @@ -3709,26 +3712,26 @@ emit_remainder_ok_encodeBlockAsm12B: JB three_bytes_emit_remainder_encodeBlockAsm12B three_bytes_emit_remainder_encodeBlockAsm12B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBlockAsm12B two_bytes_emit_remainder_encodeBlockAsm12B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBlockAsm12B JMP memmove_long_emit_remainder_encodeBlockAsm12B one_byte_emit_remainder_encodeBlockAsm12B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -3744,73 +3747,73 @@ memmove_emit_remainder_encodeBlockAsm12B: JMP emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm12B emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBlockAsm12B emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm12B emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm12B emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm12B emit_lit_memmove_emit_remainder_encodeBlockAsm12B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBlockAsm12B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBlockAsm12B memmove_long_emit_remainder_encodeBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBlockAsm12Blarge_big_loop_back: MOVOU (SI), X4 @@ -3824,967 +3827,968 @@ emit_lit_memmove_long_emit_remainder_encodeBlockAsm12Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBlockAsm12B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBlockAsm10B(dst []byte, src []byte) int +// func encodeBlockAsm10B(dst []byte, src []byte, tmp *[4096]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBlockAsm10B(SB), $4120-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000020, CX - LEAQ 24(SP), DX +TEXT ·encodeBlockAsm10B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000020, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBlockAsm10B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBlockAsm10B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeBlockAsm10B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBlockAsm10B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x9e3779b1, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x9e3779b1, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x20, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x36, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x20, R11 + IMULQ R9, R11 + SHRQ $0x36, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x20, R10 + IMULQ R9, R10 + SHRQ $0x36, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeBlockAsm10B - LEAL 1(CX), SI - MOVL 12(SP), DI - MOVL SI, BX - SUBL 16(SP), BX + LEAL 1(DX), DI + MOVL 12(SP), R8 + MOVL DI, SI + SUBL 16(SP), SI JZ repeat_extend_back_end_encodeBlockAsm10B repeat_extend_back_loop_encodeBlockAsm10B: - CMPL SI, DI + CMPL DI, R8 JBE repeat_extend_back_end_encodeBlockAsm10B - MOVB -1(DX)(BX*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(SI*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeBlockAsm10B - LEAL -1(SI), SI - DECL BX + LEAL -1(DI), DI + DECL SI JNZ repeat_extend_back_loop_encodeBlockAsm10B repeat_extend_back_end_encodeBlockAsm10B: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeBlockAsm10B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeBlockAsm10B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeBlockAsm10B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeBlockAsm10B JB three_bytes_repeat_emit_encodeBlockAsm10B three_bytes_repeat_emit_encodeBlockAsm10B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeBlockAsm10B two_bytes_repeat_emit_encodeBlockAsm10B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeBlockAsm10B JMP memmove_long_repeat_emit_encodeBlockAsm10B one_byte_repeat_emit_encodeBlockAsm10B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_repeat_emit_encodeBlockAsm10B emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm10B emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm10B emit_lit_memmove_repeat_emit_encodeBlockAsm10B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_repeat_emit_encodeBlockAsm10B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeBlockAsm10B memmove_long_repeat_emit_encodeBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R11 - SHRQ $0x05, R11 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R12 - SUBQ R10, R12 - DECQ R11 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R12 + SHRQ $0x05, R12 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R13 + SUBQ R11, R13 + DECQ R12 JA emit_lit_memmove_long_repeat_emit_encodeBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R12*1), R10 - LEAQ -32(AX)(R12*1), R13 + LEAQ -32(R10)(R13*1), R11 + LEAQ -32(CX)(R13*1), R14 emit_lit_memmove_long_repeat_emit_encodeBlockAsm10Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R13) - MOVOA X5, 16(R13) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R14) + MOVOA X5, 16(R14) + ADDQ $0x20, R14 + ADDQ $0x20, R11 ADDQ $0x20, R13 - ADDQ $0x20, R10 - ADDQ $0x20, R12 - DECQ R11 + DECQ R12 JNA emit_lit_memmove_long_repeat_emit_encodeBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R12*1), X4 - MOVOU -16(R9)(R12*1), X5 - MOVOA X4, -32(AX)(R12*1) - MOVOA X5, -16(AX)(R12*1) - ADDQ $0x20, R12 - CMPQ R8, R12 + MOVOU -32(R10)(R13*1), X4 + MOVOU -16(R10)(R13*1), X5 + MOVOA X4, -32(CX)(R13*1) + MOVOA X5, -16(CX)(R13*1) + ADDQ $0x20, R13 + CMPQ R9, R13 JAE emit_lit_memmove_long_repeat_emit_encodeBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeBlockAsm10B: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), R8 - SUBL CX, R8 - LEAQ (DX)(CX*1), R9 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R9 + SUBL DX, R9 + LEAQ (BX)(DX*1), R10 + LEAQ (BX)(SI*1), SI // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_repeat_extend_encodeBlockAsm10B: - CMPL R8, $0x10 + CMPL R9, $0x10 JB matchlen_match8_repeat_extend_encodeBlockAsm10B - MOVQ (R9)(R11*1), R10 - MOVQ 8(R9)(R11*1), R12 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + MOVQ 8(R10)(R12*1), R13 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm10B - XORQ 8(BX)(R11*1), R12 + XORQ 8(SI)(R12*1), R13 JNZ matchlen_bsf_16repeat_extend_encodeBlockAsm10B - LEAL -16(R8), R8 - LEAL 16(R11), R11 + LEAL -16(R9), R9 + LEAL 16(R12), R12 JMP matchlen_loopback_16_repeat_extend_encodeBlockAsm10B matchlen_bsf_16repeat_extend_encodeBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm10B matchlen_match8_repeat_extend_encodeBlockAsm10B: - CMPL R8, $0x08 + CMPL R9, $0x08 JB matchlen_match4_repeat_extend_encodeBlockAsm10B - MOVQ (R9)(R11*1), R10 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm10B - LEAL -8(R8), R8 - LEAL 8(R11), R11 + LEAL -8(R9), R9 + LEAL 8(R12), R12 JMP matchlen_match4_repeat_extend_encodeBlockAsm10B matchlen_bsf_8_repeat_extend_encodeBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm10B matchlen_match4_repeat_extend_encodeBlockAsm10B: - CMPL R8, $0x04 + CMPL R9, $0x04 JB matchlen_match2_repeat_extend_encodeBlockAsm10B - MOVL (R9)(R11*1), R10 - CMPL (BX)(R11*1), R10 + MOVL (R10)(R12*1), R11 + CMPL (SI)(R12*1), R11 JNE matchlen_match2_repeat_extend_encodeBlockAsm10B - LEAL -4(R8), R8 - LEAL 4(R11), R11 + LEAL -4(R9), R9 + LEAL 4(R12), R12 matchlen_match2_repeat_extend_encodeBlockAsm10B: - CMPL R8, $0x01 + CMPL R9, $0x01 JE matchlen_match1_repeat_extend_encodeBlockAsm10B JB repeat_extend_forward_end_encodeBlockAsm10B - MOVW (R9)(R11*1), R10 - CMPW (BX)(R11*1), R10 + MOVW (R10)(R12*1), R11 + CMPW (SI)(R12*1), R11 JNE matchlen_match1_repeat_extend_encodeBlockAsm10B - LEAL 2(R11), R11 - SUBL $0x02, R8 + LEAL 2(R12), R12 + SUBL $0x02, R9 JZ repeat_extend_forward_end_encodeBlockAsm10B matchlen_match1_repeat_extend_encodeBlockAsm10B: - MOVB (R9)(R11*1), R10 - CMPB (BX)(R11*1), R10 + MOVB (R10)(R12*1), R11 + CMPB (SI)(R12*1), R11 JNE repeat_extend_forward_end_encodeBlockAsm10B - LEAL 1(R11), R11 + LEAL 1(R12), R12 repeat_extend_forward_end_encodeBlockAsm10B: - ADDL R11, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI - TESTL DI, DI + ADDL R12, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI + TESTL R8, R8 JZ repeat_as_copy_encodeBlockAsm10B // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_match_repeat_encodeBlockAsm10B - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_match_repeat_encodeBlockAsm10B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_match_repeat_encodeBlockAsm10B cant_repeat_two_offset_match_repeat_encodeBlockAsm10B: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_match_repeat_encodeBlockAsm10B - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_three_match_repeat_encodeBlockAsm10B: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_two_match_repeat_encodeBlockAsm10B: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_two_offset_match_repeat_encodeBlockAsm10B: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_as_copy_encodeBlockAsm10B: // emitCopy - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeBlockAsm10B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE long_offset_short_repeat_as_copy_encodeBlockAsm10B - MOVL $0x00000001, DI - LEAL 16(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX - SUBL $0x08, BX + MOVL $0x00000001, R8 + LEAL 16(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX + SUBL $0x08, SI // emitRepeat - LEAL -4(BX), BX + LEAL -4(SI), SI JMP cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_three_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_two_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short_2b: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B long_offset_short_repeat_as_copy_encodeBlockAsm10B: - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX // emitRepeat - MOVL BX, DI - LEAL -4(BX), BX - CMPL DI, $0x08 + MOVL SI, R8 + LEAL -4(SI), SI + CMPL R8, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm10B_emit_copy_short - CMPL DI, $0x0c + CMPL R8, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JB repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm10B_emit_copy_short - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_three_repeat_as_copy_encodeBlockAsm10B_emit_copy_short: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_two_repeat_as_copy_encodeBlockAsm10B_emit_copy_short: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B repeat_two_offset_repeat_as_copy_encodeBlockAsm10B_emit_copy_short: - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B two_byte_offset_short_repeat_as_copy_encodeBlockAsm10B: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeBlockAsm10B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeBlockAsm10B - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm10B emit_copy_three_repeat_as_copy_encodeBlockAsm10B: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeBlockAsm10B: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeBlockAsm10B no_repeat_found_encodeBlockAsm10B: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBlockAsm10B - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeBlockAsm10B - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeBlockAsm10B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBlockAsm10B candidate3_match_encodeBlockAsm10B: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeBlockAsm10B candidate2_match_encodeBlockAsm10B: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeBlockAsm10B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBlockAsm10B match_extend_back_loop_encodeBlockAsm10B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBlockAsm10B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBlockAsm10B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBlockAsm10B JMP match_extend_back_loop_encodeBlockAsm10B match_extend_back_end_encodeBlockAsm10B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBlockAsm10B: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeBlockAsm10B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeBlockAsm10B - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeBlockAsm10B JB three_bytes_match_emit_encodeBlockAsm10B three_bytes_match_emit_encodeBlockAsm10B: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBlockAsm10B two_bytes_match_emit_encodeBlockAsm10B: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeBlockAsm10B JMP memmove_long_match_emit_encodeBlockAsm10B one_byte_match_emit_encodeBlockAsm10B: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBlockAsm10B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeBlockAsm10B emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm10B emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm10B emit_lit_memmove_match_emit_encodeBlockAsm10B_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBlockAsm10B: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeBlockAsm10B memmove_long_match_emit_encodeBlockAsm10B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeBlockAsm10Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeBlockAsm10B: match_nolit_loop_encodeBlockAsm10B: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeBlockAsm10B: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeBlockAsm10B - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm10B - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeBlockAsm10B - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeBlockAsm10B matchlen_bsf_16match_nolit_encodeBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeBlockAsm10B matchlen_match8_match_nolit_encodeBlockAsm10B: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeBlockAsm10B - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm10B - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeBlockAsm10B matchlen_bsf_8_match_nolit_encodeBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeBlockAsm10B matchlen_match4_match_nolit_encodeBlockAsm10B: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeBlockAsm10B - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeBlockAsm10B - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeBlockAsm10B: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeBlockAsm10B JB match_nolit_end_encodeBlockAsm10B - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeBlockAsm10B - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeBlockAsm10B matchlen_match1_match_nolit_encodeBlockAsm10B: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeBlockAsm10B - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeBlockAsm10B: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeBlockAsm10B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE long_offset_short_match_nolit_encodeBlockAsm10B - MOVL $0x00000001, SI - LEAL 16(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX - SUBL $0x08, R9 + MOVL $0x00000001, DI + LEAL 16(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R10 // emitRepeat - LEAL -4(R9), R9 + LEAL -4(R10), R10 JMP cant_repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short_2b - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm10B_emit_copy_short_2b - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short_2b - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short_2b: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm10B_emit_copy_short_2b - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B repeat_three_match_nolit_encodeBlockAsm10B_emit_copy_short_2b: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B repeat_two_match_nolit_encodeBlockAsm10B_emit_copy_short_2b: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short_2b: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B long_offset_short_match_nolit_encodeBlockAsm10B: - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX // emitRepeat - MOVL R9, SI - LEAL -4(R9), R9 - CMPL SI, $0x08 + MOVL R10, DI + LEAL -4(R10), R10 + CMPL DI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm10B_emit_copy_short - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JB repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm10B_emit_copy_short - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B repeat_three_match_nolit_encodeBlockAsm10B_emit_copy_short: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B repeat_two_match_nolit_encodeBlockAsm10B_emit_copy_short: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B repeat_two_offset_match_nolit_encodeBlockAsm10B_emit_copy_short: - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B two_byte_offset_short_match_nolit_encodeBlockAsm10B: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeBlockAsm10B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeBlockAsm10B - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm10B emit_copy_three_match_nolit_encodeBlockAsm10B: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeBlockAsm10B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBlockAsm10B - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBlockAsm10B: - MOVQ $0x9e3779b1, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x20, DI - IMULQ R8, DI - SHRQ $0x36, DI - SHLQ $0x20, BX - IMULQ R8, BX - SHRQ $0x36, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x9e3779b1, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x20, R8 + IMULQ R9, R8 + SHRQ $0x36, R8 + SHLQ $0x20, SI + IMULQ R9, SI + SHRQ $0x36, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeBlockAsm10B - INCL CX + INCL DX JMP search_loop_encodeBlockAsm10B emit_remainder_encodeBlockAsm10B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBlockAsm10B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBlockAsm10B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBlockAsm10B @@ -4793,26 +4797,26 @@ emit_remainder_ok_encodeBlockAsm10B: JB three_bytes_emit_remainder_encodeBlockAsm10B three_bytes_emit_remainder_encodeBlockAsm10B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBlockAsm10B two_bytes_emit_remainder_encodeBlockAsm10B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBlockAsm10B JMP memmove_long_emit_remainder_encodeBlockAsm10B one_byte_emit_remainder_encodeBlockAsm10B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -4828,73 +4832,73 @@ memmove_emit_remainder_encodeBlockAsm10B: JMP emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm10B emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBlockAsm10B emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm10B emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm10B emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm10B emit_lit_memmove_emit_remainder_encodeBlockAsm10B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBlockAsm10B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBlockAsm10B memmove_long_emit_remainder_encodeBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBlockAsm10Blarge_big_loop_back: MOVOU (SI), X4 @@ -4908,943 +4912,944 @@ emit_lit_memmove_long_emit_remainder_encodeBlockAsm10Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBlockAsm10B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBlockAsm8B(dst []byte, src []byte) int +// func encodeBlockAsm8B(dst []byte, src []byte, tmp *[1024]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBlockAsm8B(SB), $1048-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000008, CX - LEAQ 24(SP), DX +TEXT ·encodeBlockAsm8B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000008, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBlockAsm8B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBlockAsm8B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeBlockAsm8B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x04, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x04, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBlockAsm8B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x9e3779b1, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x38, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x9e3779b1, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x20, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x38, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x38, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 - JNE no_repeat_found_encodeBlockAsm8B - LEAL 1(CX), SI - MOVL 12(SP), DI - MOVL SI, BX - SUBL 16(SP), BX + SHLQ $0x20, R11 + IMULQ R9, R11 + SHRQ $0x38, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x20, R10 + IMULQ R9, R10 + SHRQ $0x38, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 + JNE no_repeat_found_encodeBlockAsm8B + LEAL 1(DX), DI + MOVL 12(SP), R8 + MOVL DI, SI + SUBL 16(SP), SI JZ repeat_extend_back_end_encodeBlockAsm8B repeat_extend_back_loop_encodeBlockAsm8B: - CMPL SI, DI + CMPL DI, R8 JBE repeat_extend_back_end_encodeBlockAsm8B - MOVB -1(DX)(BX*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(SI*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeBlockAsm8B - LEAL -1(SI), SI - DECL BX + LEAL -1(DI), DI + DECL SI JNZ repeat_extend_back_loop_encodeBlockAsm8B repeat_extend_back_end_encodeBlockAsm8B: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeBlockAsm8B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeBlockAsm8B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeBlockAsm8B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeBlockAsm8B JB three_bytes_repeat_emit_encodeBlockAsm8B three_bytes_repeat_emit_encodeBlockAsm8B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeBlockAsm8B two_bytes_repeat_emit_encodeBlockAsm8B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeBlockAsm8B JMP memmove_long_repeat_emit_encodeBlockAsm8B one_byte_repeat_emit_encodeBlockAsm8B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeBlockAsm8B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_repeat_emit_encodeBlockAsm8B emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm8B emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_repeat_emit_encodeBlockAsm8B emit_lit_memmove_repeat_emit_encodeBlockAsm8B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_repeat_emit_encodeBlockAsm8B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeBlockAsm8B memmove_long_repeat_emit_encodeBlockAsm8B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R11 - SHRQ $0x05, R11 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R12 - SUBQ R10, R12 - DECQ R11 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R12 + SHRQ $0x05, R12 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R13 + SUBQ R11, R13 + DECQ R12 JA emit_lit_memmove_long_repeat_emit_encodeBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R12*1), R10 - LEAQ -32(AX)(R12*1), R13 + LEAQ -32(R10)(R13*1), R11 + LEAQ -32(CX)(R13*1), R14 emit_lit_memmove_long_repeat_emit_encodeBlockAsm8Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R13) - MOVOA X5, 16(R13) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R14) + MOVOA X5, 16(R14) + ADDQ $0x20, R14 + ADDQ $0x20, R11 ADDQ $0x20, R13 - ADDQ $0x20, R10 - ADDQ $0x20, R12 - DECQ R11 + DECQ R12 JNA emit_lit_memmove_long_repeat_emit_encodeBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R12*1), X4 - MOVOU -16(R9)(R12*1), X5 - MOVOA X4, -32(AX)(R12*1) - MOVOA X5, -16(AX)(R12*1) - ADDQ $0x20, R12 - CMPQ R8, R12 + MOVOU -32(R10)(R13*1), X4 + MOVOU -16(R10)(R13*1), X5 + MOVOA X4, -32(CX)(R13*1) + MOVOA X5, -16(CX)(R13*1) + ADDQ $0x20, R13 + CMPQ R9, R13 JAE emit_lit_memmove_long_repeat_emit_encodeBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeBlockAsm8B: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), R8 - SUBL CX, R8 - LEAQ (DX)(CX*1), R9 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R9 + SUBL DX, R9 + LEAQ (BX)(DX*1), R10 + LEAQ (BX)(SI*1), SI // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_repeat_extend_encodeBlockAsm8B: - CMPL R8, $0x10 + CMPL R9, $0x10 JB matchlen_match8_repeat_extend_encodeBlockAsm8B - MOVQ (R9)(R11*1), R10 - MOVQ 8(R9)(R11*1), R12 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + MOVQ 8(R10)(R12*1), R13 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm8B - XORQ 8(BX)(R11*1), R12 + XORQ 8(SI)(R12*1), R13 JNZ matchlen_bsf_16repeat_extend_encodeBlockAsm8B - LEAL -16(R8), R8 - LEAL 16(R11), R11 + LEAL -16(R9), R9 + LEAL 16(R12), R12 JMP matchlen_loopback_16_repeat_extend_encodeBlockAsm8B matchlen_bsf_16repeat_extend_encodeBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm8B matchlen_match8_repeat_extend_encodeBlockAsm8B: - CMPL R8, $0x08 + CMPL R9, $0x08 JB matchlen_match4_repeat_extend_encodeBlockAsm8B - MOVQ (R9)(R11*1), R10 - XORQ (BX)(R11*1), R10 + MOVQ (R10)(R12*1), R11 + XORQ (SI)(R12*1), R11 JNZ matchlen_bsf_8_repeat_extend_encodeBlockAsm8B - LEAL -8(R8), R8 - LEAL 8(R11), R11 + LEAL -8(R9), R9 + LEAL 8(R12), R12 JMP matchlen_match4_repeat_extend_encodeBlockAsm8B matchlen_bsf_8_repeat_extend_encodeBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP repeat_extend_forward_end_encodeBlockAsm8B matchlen_match4_repeat_extend_encodeBlockAsm8B: - CMPL R8, $0x04 + CMPL R9, $0x04 JB matchlen_match2_repeat_extend_encodeBlockAsm8B - MOVL (R9)(R11*1), R10 - CMPL (BX)(R11*1), R10 + MOVL (R10)(R12*1), R11 + CMPL (SI)(R12*1), R11 JNE matchlen_match2_repeat_extend_encodeBlockAsm8B - LEAL -4(R8), R8 - LEAL 4(R11), R11 + LEAL -4(R9), R9 + LEAL 4(R12), R12 matchlen_match2_repeat_extend_encodeBlockAsm8B: - CMPL R8, $0x01 + CMPL R9, $0x01 JE matchlen_match1_repeat_extend_encodeBlockAsm8B JB repeat_extend_forward_end_encodeBlockAsm8B - MOVW (R9)(R11*1), R10 - CMPW (BX)(R11*1), R10 + MOVW (R10)(R12*1), R11 + CMPW (SI)(R12*1), R11 JNE matchlen_match1_repeat_extend_encodeBlockAsm8B - LEAL 2(R11), R11 - SUBL $0x02, R8 + LEAL 2(R12), R12 + SUBL $0x02, R9 JZ repeat_extend_forward_end_encodeBlockAsm8B matchlen_match1_repeat_extend_encodeBlockAsm8B: - MOVB (R9)(R11*1), R10 - CMPB (BX)(R11*1), R10 + MOVB (R10)(R12*1), R11 + CMPB (SI)(R12*1), R11 JNE repeat_extend_forward_end_encodeBlockAsm8B - LEAL 1(R11), R11 + LEAL 1(R12), R12 repeat_extend_forward_end_encodeBlockAsm8B: - ADDL R11, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI - TESTL DI, DI + ADDL R12, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI + TESTL R8, R8 JZ repeat_as_copy_encodeBlockAsm8B // emitRepeat - MOVL BX, SI - LEAL -4(BX), BX - CMPL SI, $0x08 + MOVL SI, DI + LEAL -4(SI), SI + CMPL DI, $0x08 JBE repeat_two_match_repeat_encodeBlockAsm8B - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_match_repeat_encodeBlockAsm8B cant_repeat_two_offset_match_repeat_encodeBlockAsm8B: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_match_repeat_encodeBlockAsm8B - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_three_match_repeat_encodeBlockAsm8B: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_two_match_repeat_encodeBlockAsm8B: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_as_copy_encodeBlockAsm8B: // emitCopy - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeBlockAsm8B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE long_offset_short_repeat_as_copy_encodeBlockAsm8B - MOVL $0x00000001, DI - LEAL 16(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX - SUBL $0x08, BX + MOVL $0x00000001, R8 + LEAL 16(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX + SUBL $0x08, SI // emitRepeat - LEAL -4(BX), BX + LEAL -4(SI), SI JMP cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b - MOVL BX, SI - LEAL -4(BX), BX - CMPL SI, $0x08 + MOVL SI, DI + LEAL -4(SI), SI + CMPL DI, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_three_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_two_repeat_as_copy_encodeBlockAsm8B_emit_copy_short_2b: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B long_offset_short_repeat_as_copy_encodeBlockAsm8B: - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX // emitRepeat - MOVL BX, SI - LEAL -4(BX), BX - CMPL SI, $0x08 + MOVL SI, DI + LEAL -4(SI), SI + CMPL DI, $0x08 JBE repeat_two_repeat_as_copy_encodeBlockAsm8B_emit_copy_short - CMPL SI, $0x0c + CMPL DI, $0x0c JAE cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm8B_emit_copy_short cant_repeat_two_offset_repeat_as_copy_encodeBlockAsm8B_emit_copy_short: - CMPL BX, $0x00000104 + CMPL SI, $0x00000104 JB repeat_three_repeat_as_copy_encodeBlockAsm8B_emit_copy_short - LEAL -256(BX), BX - MOVW $0x0019, (AX) - MOVW BX, 2(AX) - ADDQ $0x04, AX + LEAL -256(SI), SI + MOVW $0x0019, (CX) + MOVW SI, 2(CX) + ADDQ $0x04, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_three_repeat_as_copy_encodeBlockAsm8B_emit_copy_short: - LEAL -4(BX), BX - MOVW $0x0015, (AX) - MOVB BL, 2(AX) - ADDQ $0x03, AX + LEAL -4(SI), SI + MOVW $0x0015, (CX) + MOVB SI, 2(CX) + ADDQ $0x03, CX JMP repeat_end_emit_encodeBlockAsm8B repeat_two_repeat_as_copy_encodeBlockAsm8B_emit_copy_short: - SHLL $0x02, BX - ORL $0x01, BX - MOVW BX, (AX) - ADDQ $0x02, AX + SHLL $0x02, SI + ORL $0x01, SI + MOVW SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B - XORQ DI, DI - LEAL 1(DI)(BX*4), BX - MOVB SI, 1(AX) - SARL $0x08, SI - SHLL $0x05, SI - ORL SI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + XORQ R8, R8 + LEAL 1(R8)(SI*4), SI + MOVB DI, 1(CX) + SARL $0x08, DI + SHLL $0x05, DI + ORL DI, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B two_byte_offset_short_repeat_as_copy_encodeBlockAsm8B: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeBlockAsm8B - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeBlockAsm8B emit_copy_three_repeat_as_copy_encodeBlockAsm8B: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeBlockAsm8B: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeBlockAsm8B no_repeat_found_encodeBlockAsm8B: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBlockAsm8B - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeBlockAsm8B - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeBlockAsm8B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBlockAsm8B candidate3_match_encodeBlockAsm8B: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeBlockAsm8B candidate2_match_encodeBlockAsm8B: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeBlockAsm8B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBlockAsm8B match_extend_back_loop_encodeBlockAsm8B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBlockAsm8B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBlockAsm8B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBlockAsm8B JMP match_extend_back_loop_encodeBlockAsm8B match_extend_back_end_encodeBlockAsm8B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBlockAsm8B: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeBlockAsm8B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeBlockAsm8B - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeBlockAsm8B JB three_bytes_match_emit_encodeBlockAsm8B three_bytes_match_emit_encodeBlockAsm8B: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBlockAsm8B two_bytes_match_emit_encodeBlockAsm8B: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeBlockAsm8B JMP memmove_long_match_emit_encodeBlockAsm8B one_byte_match_emit_encodeBlockAsm8B: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBlockAsm8B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeBlockAsm8B emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm8B emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBlockAsm8B emit_lit_memmove_match_emit_encodeBlockAsm8B_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBlockAsm8B: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeBlockAsm8B memmove_long_match_emit_encodeBlockAsm8B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeBlockAsm8Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) - ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 + ADDQ $0x20, R12 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeBlockAsm8B: match_nolit_loop_encodeBlockAsm8B: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeBlockAsm8B: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeBlockAsm8B - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm8B - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeBlockAsm8B - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeBlockAsm8B matchlen_bsf_16match_nolit_encodeBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeBlockAsm8B matchlen_match8_match_nolit_encodeBlockAsm8B: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeBlockAsm8B - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeBlockAsm8B - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeBlockAsm8B matchlen_bsf_8_match_nolit_encodeBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeBlockAsm8B matchlen_match4_match_nolit_encodeBlockAsm8B: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeBlockAsm8B - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeBlockAsm8B - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeBlockAsm8B: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeBlockAsm8B JB match_nolit_end_encodeBlockAsm8B - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeBlockAsm8B - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeBlockAsm8B matchlen_match1_match_nolit_encodeBlockAsm8B: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeBlockAsm8B - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeBlockAsm8B: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeBlockAsm8B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE long_offset_short_match_nolit_encodeBlockAsm8B - MOVL $0x00000001, SI - LEAL 16(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX - SUBL $0x08, R9 + MOVL $0x00000001, DI + LEAL 16(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R10 // emitRepeat - LEAL -4(R9), R9 + LEAL -4(R10), R10 JMP cant_repeat_two_offset_match_nolit_encodeBlockAsm8B_emit_copy_short_2b - MOVL R9, BX - LEAL -4(R9), R9 - CMPL BX, $0x08 + MOVL R10, SI + LEAL -4(R10), R10 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm8B_emit_copy_short_2b - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm8B_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBlockAsm8B_emit_copy_short_2b: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm8B_emit_copy_short_2b - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B repeat_three_match_nolit_encodeBlockAsm8B_emit_copy_short_2b: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B repeat_two_match_nolit_encodeBlockAsm8B_emit_copy_short_2b: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B long_offset_short_match_nolit_encodeBlockAsm8B: - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX // emitRepeat - MOVL R9, BX - LEAL -4(R9), R9 - CMPL BX, $0x08 + MOVL R10, SI + LEAL -4(R10), R10 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBlockAsm8B_emit_copy_short - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBlockAsm8B_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBlockAsm8B_emit_copy_short: - CMPL R9, $0x00000104 + CMPL R10, $0x00000104 JB repeat_three_match_nolit_encodeBlockAsm8B_emit_copy_short - LEAL -256(R9), R9 - MOVW $0x0019, (AX) - MOVW R9, 2(AX) - ADDQ $0x04, AX + LEAL -256(R10), R10 + MOVW $0x0019, (CX) + MOVW R10, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B repeat_three_match_nolit_encodeBlockAsm8B_emit_copy_short: - LEAL -4(R9), R9 - MOVW $0x0015, (AX) - MOVB R9, 2(AX) - ADDQ $0x03, AX + LEAL -4(R10), R10 + MOVW $0x0015, (CX) + MOVB R10, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B repeat_two_match_nolit_encodeBlockAsm8B_emit_copy_short: - SHLL $0x02, R9 - ORL $0x01, R9 - MOVW R9, (AX) - ADDQ $0x02, AX + SHLL $0x02, R10 + ORL $0x01, R10 + MOVW R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B - XORQ SI, SI - LEAL 1(SI)(R9*4), R9 - MOVB BL, 1(AX) - SARL $0x08, BX - SHLL $0x05, BX - ORL BX, R9 - MOVB R9, (AX) - ADDQ $0x02, AX + XORQ DI, DI + LEAL 1(DI)(R10*4), R10 + MOVB SI, 1(CX) + SARL $0x08, SI + SHLL $0x05, SI + ORL SI, R10 + MOVB R10, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B two_byte_offset_short_match_nolit_encodeBlockAsm8B: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeBlockAsm8B - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBlockAsm8B emit_copy_three_match_nolit_encodeBlockAsm8B: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeBlockAsm8B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBlockAsm8B - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBlockAsm8B: - MOVQ $0x9e3779b1, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x20, DI - IMULQ R8, DI - SHRQ $0x38, DI - SHLQ $0x20, BX - IMULQ R8, BX - SHRQ $0x38, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x9e3779b1, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x20, R8 + IMULQ R9, R8 + SHRQ $0x38, R8 + SHLQ $0x20, SI + IMULQ R9, SI + SHRQ $0x38, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeBlockAsm8B - INCL CX + INCL DX JMP search_loop_encodeBlockAsm8B emit_remainder_encodeBlockAsm8B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBlockAsm8B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBlockAsm8B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBlockAsm8B @@ -5853,26 +5858,26 @@ emit_remainder_ok_encodeBlockAsm8B: JB three_bytes_emit_remainder_encodeBlockAsm8B three_bytes_emit_remainder_encodeBlockAsm8B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBlockAsm8B two_bytes_emit_remainder_encodeBlockAsm8B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBlockAsm8B JMP memmove_long_emit_remainder_encodeBlockAsm8B one_byte_emit_remainder_encodeBlockAsm8B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -5888,73 +5893,73 @@ memmove_emit_remainder_encodeBlockAsm8B: JMP emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm8B emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBlockAsm8B emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm8B emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm8B emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBlockAsm8B emit_lit_memmove_emit_remainder_encodeBlockAsm8B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBlockAsm8B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBlockAsm8B memmove_long_emit_remainder_encodeBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBlockAsm8Blarge_big_loop_back: MOVOU (SI), X4 @@ -5968,961 +5973,962 @@ emit_lit_memmove_long_emit_remainder_encodeBlockAsm8Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBlockAsm8B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBetterBlockAsm(dst []byte, src []byte) int +// func encodeBetterBlockAsm(dst []byte, src []byte, tmp *[589824]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBetterBlockAsm(SB), $589848-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00001200, CX - LEAQ 24(SP), DX +TEXT ·encodeBetterBlockAsm(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00001200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBetterBlockAsm: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBetterBlockAsm MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -6(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -6(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeBetterBlockAsm: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x07, BX - CMPL BX, $0x63 + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x07, SI + CMPL SI, $0x63 JBE check_maxskip_ok_encodeBetterBlockAsm - LEAL 100(CX), BX + LEAL 100(DX), SI JMP check_maxskip_cont_encodeBetterBlockAsm check_maxskip_ok_encodeBetterBlockAsm: - LEAL 1(CX)(BX*1), BX + LEAL 1(DX)(SI*1), SI check_maxskip_cont_encodeBetterBlockAsm: - CMPL BX, 8(SP) + CMPL SI, 8(SP) JAE emit_remainder_encodeBetterBlockAsm - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x00cf1bbcdcbfa563, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x2f, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 524312(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 524312(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x00cf1bbcdcbfa563, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x2f, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL 524288(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 524288(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeBetterBlockAsm - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeBetterBlockAsm - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeBetterBlockAsm no_short_found_encodeBetterBlockAsm: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeBetterBlockAsm - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeBetterBlockAsm - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBetterBlockAsm candidateS_match_encodeBetterBlockAsm: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x2f, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x2f, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBetterBlockAsm - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeBetterBlockAsm: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBetterBlockAsm match_extend_back_loop_encodeBetterBlockAsm: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBetterBlockAsm - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBetterBlockAsm - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBetterBlockAsm JMP match_extend_back_loop_encodeBetterBlockAsm match_extend_back_end_encodeBetterBlockAsm: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 5(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 5(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBetterBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBetterBlockAsm: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeBetterBlockAsm: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeBetterBlockAsm - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeBetterBlockAsm - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeBetterBlockAsm matchlen_bsf_16match_nolit_encodeBetterBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeBetterBlockAsm matchlen_match8_match_nolit_encodeBetterBlockAsm: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeBetterBlockAsm - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeBetterBlockAsm matchlen_bsf_8_match_nolit_encodeBetterBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeBetterBlockAsm matchlen_match4_match_nolit_encodeBetterBlockAsm: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeBetterBlockAsm - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeBetterBlockAsm - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeBetterBlockAsm: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeBetterBlockAsm JB match_nolit_end_encodeBetterBlockAsm - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeBetterBlockAsm - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeBetterBlockAsm matchlen_match1_match_nolit_encodeBetterBlockAsm: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeBetterBlockAsm - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeBetterBlockAsm: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - CMPL 16(SP), DI + CMPL 16(SP), R8 JEQ match_is_repeat_encodeBetterBlockAsm - CMPL R11, $0x01 + CMPL R12, $0x01 JA match_length_ok_encodeBetterBlockAsm - CMPL DI, $0x0000ffff + CMPL R8, $0x0000ffff JBE match_length_ok_encodeBetterBlockAsm - MOVL 20(SP), CX - INCL CX + MOVL 20(SP), DX + INCL DX JMP search_loop_encodeBetterBlockAsm match_length_ok_encodeBetterBlockAsm: - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeBetterBlockAsm - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeBetterBlockAsm - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeBetterBlockAsm - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_match_emit_encodeBetterBlockAsm - CMPL BX, $0x01000000 + CMPL SI, $0x01000000 JB four_bytes_match_emit_encodeBetterBlockAsm - MOVB $0xfc, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP memmove_long_match_emit_encodeBetterBlockAsm four_bytes_match_emit_encodeBetterBlockAsm: - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_encodeBetterBlockAsm three_bytes_match_emit_encodeBetterBlockAsm: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBetterBlockAsm two_bytes_match_emit_encodeBetterBlockAsm: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeBetterBlockAsm JMP memmove_long_match_emit_encodeBetterBlockAsm one_byte_match_emit_encodeBetterBlockAsm: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBetterBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm emit_lit_memmove_match_emit_encodeBetterBlockAsm_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBetterBlockAsm: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeBetterBlockAsm memmove_long_match_emit_encodeBetterBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeBetterBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeBetterBlockAsmlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeBetterBlockAsmlarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBetterBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeBetterBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeBetterBlockAsm: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy - CMPL DI, $0x00010000 + CMPL R8, $0x00010000 JB two_byte_offset_match_nolit_encodeBetterBlockAsm - CMPL R11, $0x40 + CMPL R12, $0x40 JBE four_bytes_remain_match_nolit_encodeBetterBlockAsm - MOVB $0xff, (AX) - MOVL DI, 1(AX) - LEAL -64(R11), R11 - ADDQ $0x05, AX - CMPL R11, $0x04 + MOVB $0xff, (CX) + MOVL R8, 1(CX) + LEAL -64(R12), R12 + ADDQ $0x05, CX + CMPL R12, $0x04 JB four_bytes_remain_match_nolit_encodeBetterBlockAsm // emitRepeat emit_repeat_again_match_nolit_encodeBetterBlockAsm_emit_copy: - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm_emit_copy - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm_emit_copy - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_encodeBetterBlockAsm_emit_copy - CMPL R11, $0x0100ffff + CMPL R12, $0x0100ffff JB repeat_five_match_nolit_encodeBetterBlockAsm_emit_copy - LEAL -16842747(R11), R11 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R12), R12 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_encodeBetterBlockAsm_emit_copy repeat_five_match_nolit_encodeBetterBlockAsm_emit_copy: - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_four_match_nolit_encodeBetterBlockAsm_emit_copy: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_three_match_nolit_encodeBetterBlockAsm_emit_copy: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_match_nolit_encodeBetterBlockAsm_emit_copy: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm four_bytes_remain_match_nolit_encodeBetterBlockAsm: - TESTL R11, R11 + TESTL R12, R12 JZ match_nolit_emitcopy_end_encodeBetterBlockAsm - XORL BX, BX - LEAL -1(BX)(R11*4), R11 - MOVB R11, (AX) - MOVL DI, 1(AX) - ADDQ $0x05, AX + XORL SI, SI + LEAL -1(SI)(R12*4), R12 + MOVB R12, (CX) + MOVL R8, 1(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm two_byte_offset_match_nolit_encodeBetterBlockAsm: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeBetterBlockAsm - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE long_offset_short_match_nolit_encodeBetterBlockAsm - MOVL $0x00000001, BX - LEAL 16(BX), BX - MOVB DI, 1(AX) - MOVL DI, R8 - SHRL $0x08, R8 - SHLL $0x05, R8 - ORL R8, BX - MOVB BL, (AX) - ADDQ $0x02, AX - SUBL $0x08, R11 + MOVL $0x00000001, SI + LEAL 16(SI), SI + MOVB R8, 1(CX) + MOVL R8, R9 + SHRL $0x08, R9 + SHLL $0x05, R9 + ORL R9, SI + MOVB SI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R12 // emitRepeat - LEAL -4(R11), R11 + LEAL -4(R12), R12 JMP cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b emit_repeat_again_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b - CMPL R11, $0x0100ffff + CMPL R12, $0x0100ffff JB repeat_five_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b - LEAL -16842747(R11), R11 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R12), R12 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b repeat_five_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_four_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_three_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short_2b: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm long_offset_short_match_nolit_encodeBetterBlockAsm: - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX // emitRepeat emit_repeat_again_match_nolit_encodeBetterBlockAsm_emit_copy_short: - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm_emit_copy_short - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm_emit_copy_short - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_encodeBetterBlockAsm_emit_copy_short - CMPL R11, $0x0100ffff + CMPL R12, $0x0100ffff JB repeat_five_match_nolit_encodeBetterBlockAsm_emit_copy_short - LEAL -16842747(R11), R11 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R12), R12 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_encodeBetterBlockAsm_emit_copy_short repeat_five_match_nolit_encodeBetterBlockAsm_emit_copy_short: - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_four_match_nolit_encodeBetterBlockAsm_emit_copy_short: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_three_match_nolit_encodeBetterBlockAsm_emit_copy_short: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_match_nolit_encodeBetterBlockAsm_emit_copy_short: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_offset_match_nolit_encodeBetterBlockAsm_emit_copy_short: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm two_byte_offset_short_match_nolit_encodeBetterBlockAsm: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeBetterBlockAsm - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeBetterBlockAsm - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm emit_copy_three_match_nolit_encodeBetterBlockAsm: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm match_is_repeat_encodeBetterBlockAsm: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_repeat_encodeBetterBlockAsm - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_repeat_encodeBetterBlockAsm - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_repeat_encodeBetterBlockAsm - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_match_emit_repeat_encodeBetterBlockAsm - CMPL BX, $0x01000000 + CMPL SI, $0x01000000 JB four_bytes_match_emit_repeat_encodeBetterBlockAsm - MOVB $0xfc, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm four_bytes_match_emit_repeat_encodeBetterBlockAsm: - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm three_bytes_match_emit_repeat_encodeBetterBlockAsm: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm two_bytes_match_emit_repeat_encodeBetterBlockAsm: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_repeat_encodeBetterBlockAsm JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm one_byte_match_emit_repeat_encodeBetterBlockAsm: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_repeat_encodeBetterBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_33through64 emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_repeat_encodeBetterBlockAsm memmove_long_match_emit_repeat_encodeBetterBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsmlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsmlarge_big_loop_back emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_repeat_encodeBetterBlockAsm: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitRepeat emit_repeat_again_match_nolit_repeat_encodeBetterBlockAsm: - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_repeat_encodeBetterBlockAsm - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_repeat_encodeBetterBlockAsm - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_repeat_encodeBetterBlockAsm - CMPL R11, $0x0100ffff + CMPL R12, $0x0100ffff JB repeat_five_match_nolit_repeat_encodeBetterBlockAsm - LEAL -16842747(R11), R11 - MOVL $0xfffb001d, (AX) - MOVB $0xff, 4(AX) - ADDQ $0x05, AX + LEAL -16842747(R12), R12 + MOVL $0xfffb001d, (CX) + MOVB $0xff, 4(CX) + ADDQ $0x05, CX JMP emit_repeat_again_match_nolit_repeat_encodeBetterBlockAsm repeat_five_match_nolit_repeat_encodeBetterBlockAsm: - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_four_match_nolit_repeat_encodeBetterBlockAsm: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_three_match_nolit_repeat_encodeBetterBlockAsm: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_match_nolit_repeat_encodeBetterBlockAsm: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX match_nolit_emitcopy_end_encodeBetterBlockAsm: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBetterBlockAsm - CMPQ AX, (SP) - JB match_nolit_dst_ok_encodeBetterBlockAsm - MOVQ $0x00000000, ret+48(FP) - RET - -match_nolit_dst_ok_encodeBetterBlockAsm: - MOVQ $0x00cf1bbcdcbfa563, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x2f, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x32, R10 - SHLQ $0x08, R11 - IMULQ BX, R11 - SHRQ $0x2f, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x32, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 524312(SP)(R10*4) - MOVL R13, 524312(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + CMPQ CX, (SP) + JB match_nolit_dst_ok_encodeBetterBlockAsm + MOVQ $0x00000000, ret+56(FP) + RET + +match_nolit_dst_ok_encodeBetterBlockAsm: + MOVQ $0x00cf1bbcdcbfa563, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x08, R10 + IMULQ SI, R10 + SHRQ $0x2f, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x32, R11 + SHLQ $0x08, R12 + IMULQ SI, R12 + SHRQ $0x2f, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x32, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 524288(AX)(R11*4) + MOVL R14, 524288(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeBetterBlockAsm: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeBetterBlockAsm - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x2f, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x08, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x2f, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x08, R11 + IMULQ SI, R11 + SHRQ $0x2f, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeBetterBlockAsm emit_remainder_encodeBetterBlockAsm: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 5(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 5(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBetterBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBetterBlockAsm: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBetterBlockAsm - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBetterBlockAsm @@ -6932,41 +6938,41 @@ emit_remainder_ok_encodeBetterBlockAsm: JB three_bytes_emit_remainder_encodeBetterBlockAsm CMPL DX, $0x01000000 JB four_bytes_emit_remainder_encodeBetterBlockAsm - MOVB $0xfc, (AX) - MOVL DX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL DX, 1(CX) + ADDQ $0x05, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm four_bytes_emit_remainder_encodeBetterBlockAsm: MOVL DX, BX SHRL $0x10, BX - MOVB $0xf8, (AX) - MOVW DX, 1(AX) - MOVB BL, 3(AX) - ADDQ $0x04, AX + MOVB $0xf8, (CX) + MOVW DX, 1(CX) + MOVB BL, 3(CX) + ADDQ $0x04, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm three_bytes_emit_remainder_encodeBetterBlockAsm: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm two_bytes_emit_remainder_encodeBetterBlockAsm: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBetterBlockAsm JMP memmove_long_emit_remainder_encodeBetterBlockAsm one_byte_emit_remainder_encodeBetterBlockAsm: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBetterBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -6982,73 +6988,73 @@ memmove_emit_remainder_encodeBetterBlockAsm: JMP emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm emit_lit_memmove_emit_remainder_encodeBetterBlockAsm_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBetterBlockAsm: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBetterBlockAsm memmove_long_emit_remainder_encodeBetterBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsmlarge_big_loop_back: MOVOU (SI), X4 @@ -7062,903 +7068,904 @@ emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsmlarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsmlarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBetterBlockAsm: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBetterBlockAsm4MB(dst []byte, src []byte) int +// func encodeBetterBlockAsm4MB(dst []byte, src []byte, tmp *[589824]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBetterBlockAsm4MB(SB), $589848-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00001200, CX - LEAQ 24(SP), DX +TEXT ·encodeBetterBlockAsm4MB(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00001200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBetterBlockAsm4MB: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBetterBlockAsm4MB MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -6(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -6(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeBetterBlockAsm4MB: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x07, BX - CMPL BX, $0x63 + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x07, SI + CMPL SI, $0x63 JBE check_maxskip_ok_encodeBetterBlockAsm4MB - LEAL 100(CX), BX + LEAL 100(DX), SI JMP check_maxskip_cont_encodeBetterBlockAsm4MB check_maxskip_ok_encodeBetterBlockAsm4MB: - LEAL 1(CX)(BX*1), BX + LEAL 1(DX)(SI*1), SI check_maxskip_cont_encodeBetterBlockAsm4MB: - CMPL BX, 8(SP) + CMPL SI, 8(SP) JAE emit_remainder_encodeBetterBlockAsm4MB - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x00cf1bbcdcbfa563, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x2f, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 524312(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 524312(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x00cf1bbcdcbfa563, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x2f, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL 524288(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 524288(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeBetterBlockAsm4MB - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeBetterBlockAsm4MB - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeBetterBlockAsm4MB no_short_found_encodeBetterBlockAsm4MB: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeBetterBlockAsm4MB - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeBetterBlockAsm4MB - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBetterBlockAsm4MB candidateS_match_encodeBetterBlockAsm4MB: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x2f, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x2f, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBetterBlockAsm4MB - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeBetterBlockAsm4MB: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBetterBlockAsm4MB match_extend_back_loop_encodeBetterBlockAsm4MB: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBetterBlockAsm4MB - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBetterBlockAsm4MB - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBetterBlockAsm4MB JMP match_extend_back_loop_encodeBetterBlockAsm4MB match_extend_back_end_encodeBetterBlockAsm4MB: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 4(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 4(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBetterBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBetterBlockAsm4MB: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeBetterBlockAsm4MB: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeBetterBlockAsm4MB - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm4MB - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeBetterBlockAsm4MB - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeBetterBlockAsm4MB matchlen_bsf_16match_nolit_encodeBetterBlockAsm4MB: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeBetterBlockAsm4MB matchlen_match8_match_nolit_encodeBetterBlockAsm4MB: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeBetterBlockAsm4MB - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm4MB - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeBetterBlockAsm4MB matchlen_bsf_8_match_nolit_encodeBetterBlockAsm4MB: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeBetterBlockAsm4MB matchlen_match4_match_nolit_encodeBetterBlockAsm4MB: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeBetterBlockAsm4MB - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeBetterBlockAsm4MB - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeBetterBlockAsm4MB: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeBetterBlockAsm4MB JB match_nolit_end_encodeBetterBlockAsm4MB - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeBetterBlockAsm4MB - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeBetterBlockAsm4MB matchlen_match1_match_nolit_encodeBetterBlockAsm4MB: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeBetterBlockAsm4MB - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeBetterBlockAsm4MB: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - CMPL 16(SP), DI + CMPL 16(SP), R8 JEQ match_is_repeat_encodeBetterBlockAsm4MB - CMPL R11, $0x01 + CMPL R12, $0x01 JA match_length_ok_encodeBetterBlockAsm4MB - CMPL DI, $0x0000ffff + CMPL R8, $0x0000ffff JBE match_length_ok_encodeBetterBlockAsm4MB - MOVL 20(SP), CX - INCL CX + MOVL 20(SP), DX + INCL DX JMP search_loop_encodeBetterBlockAsm4MB match_length_ok_encodeBetterBlockAsm4MB: - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeBetterBlockAsm4MB - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeBetterBlockAsm4MB - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeBetterBlockAsm4MB - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_match_emit_encodeBetterBlockAsm4MB - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_encodeBetterBlockAsm4MB three_bytes_match_emit_encodeBetterBlockAsm4MB: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBetterBlockAsm4MB two_bytes_match_emit_encodeBetterBlockAsm4MB: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeBetterBlockAsm4MB JMP memmove_long_match_emit_encodeBetterBlockAsm4MB one_byte_match_emit_encodeBetterBlockAsm4MB: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBetterBlockAsm4MB: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_encodeBetterBlockAsm4MB_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBetterBlockAsm4MB: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeBetterBlockAsm4MB memmove_long_match_emit_encodeBetterBlockAsm4MB: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeBetterBlockAsm4MBlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm4MBlarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeBetterBlockAsm4MB: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy - CMPL DI, $0x00010000 + CMPL R8, $0x00010000 JB two_byte_offset_match_nolit_encodeBetterBlockAsm4MB - CMPL R11, $0x40 + CMPL R12, $0x40 JBE four_bytes_remain_match_nolit_encodeBetterBlockAsm4MB - MOVB $0xff, (AX) - MOVL DI, 1(AX) - LEAL -64(R11), R11 - ADDQ $0x05, AX - CMPL R11, $0x04 + MOVB $0xff, (CX) + MOVL R8, 1(CX) + LEAL -64(R12), R12 + ADDQ $0x05, CX + CMPL R12, $0x04 JB four_bytes_remain_match_nolit_encodeBetterBlockAsm4MB // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm4MB_emit_copy - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm4MB_emit_copy - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_encodeBetterBlockAsm4MB_emit_copy - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_four_match_nolit_encodeBetterBlockAsm4MB_emit_copy: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_three_match_nolit_encodeBetterBlockAsm4MB_emit_copy: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_match_nolit_encodeBetterBlockAsm4MB_emit_copy: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB four_bytes_remain_match_nolit_encodeBetterBlockAsm4MB: - TESTL R11, R11 + TESTL R12, R12 JZ match_nolit_emitcopy_end_encodeBetterBlockAsm4MB - XORL BX, BX - LEAL -1(BX)(R11*4), R11 - MOVB R11, (AX) - MOVL DI, 1(AX) - ADDQ $0x05, AX + XORL SI, SI + LEAL -1(SI)(R12*4), R12 + MOVB R12, (CX) + MOVL R8, 1(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB two_byte_offset_match_nolit_encodeBetterBlockAsm4MB: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeBetterBlockAsm4MB - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE long_offset_short_match_nolit_encodeBetterBlockAsm4MB - MOVL $0x00000001, BX - LEAL 16(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX - SUBL $0x08, R11 + MOVL $0x00000001, SI + LEAL 16(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R12 // emitRepeat - LEAL -4(R11), R11 + LEAL -4(R12), R12 JMP cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_four_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_three_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short_2b: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB long_offset_short_match_nolit_encodeBetterBlockAsm4MB: - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_four_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_three_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_offset_match_nolit_encodeBetterBlockAsm4MB_emit_copy_short: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB two_byte_offset_short_match_nolit_encodeBetterBlockAsm4MB: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeBetterBlockAsm4MB - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeBetterBlockAsm4MB - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB emit_copy_three_match_nolit_encodeBetterBlockAsm4MB: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB match_is_repeat_encodeBetterBlockAsm4MB: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_repeat_encodeBetterBlockAsm4MB - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_repeat_encodeBetterBlockAsm4MB - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_repeat_encodeBetterBlockAsm4MB - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_match_emit_repeat_encodeBetterBlockAsm4MB - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm4MB three_bytes_match_emit_repeat_encodeBetterBlockAsm4MB: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm4MB two_bytes_match_emit_repeat_encodeBetterBlockAsm4MB: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_repeat_encodeBetterBlockAsm4MB JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm4MB one_byte_match_emit_repeat_encodeBetterBlockAsm4MB: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_repeat_encodeBetterBlockAsm4MB: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_17through32 JMP emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_33through64 emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm4MB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm4MB_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm4MB: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_repeat_encodeBetterBlockAsm4MB memmove_long_match_emit_repeat_encodeBetterBlockAsm4MB: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm4MBlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm4MBlarge_big_loop_back emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_repeat_encodeBetterBlockAsm4MB: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_repeat_encodeBetterBlockAsm4MB - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm4MB - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm4MB cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm4MB: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_repeat_encodeBetterBlockAsm4MB - CMPL R11, $0x00010100 + CMPL R12, $0x00010100 JB repeat_four_match_nolit_repeat_encodeBetterBlockAsm4MB - LEAL -65536(R11), R11 - MOVL R11, DI - MOVW $0x001d, (AX) - MOVW R11, 2(AX) - SARL $0x10, DI - MOVB DI, 4(AX) - ADDQ $0x05, AX + LEAL -65536(R12), R12 + MOVL R12, R8 + MOVW $0x001d, (CX) + MOVW R12, 2(CX) + SARL $0x10, R8 + MOVB R8, 4(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_four_match_nolit_repeat_encodeBetterBlockAsm4MB: - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_three_match_nolit_repeat_encodeBetterBlockAsm4MB: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_match_nolit_repeat_encodeBetterBlockAsm4MB: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm4MB repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm4MB: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX match_nolit_emitcopy_end_encodeBetterBlockAsm4MB: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBetterBlockAsm4MB - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBetterBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBetterBlockAsm4MB: - MOVQ $0x00cf1bbcdcbfa563, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x2f, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x32, R10 - SHLQ $0x08, R11 - IMULQ BX, R11 - SHRQ $0x2f, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x32, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 524312(SP)(R10*4) - MOVL R13, 524312(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x00cf1bbcdcbfa563, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x08, R10 + IMULQ SI, R10 + SHRQ $0x2f, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x32, R11 + SHLQ $0x08, R12 + IMULQ SI, R12 + SHRQ $0x2f, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x32, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 524288(AX)(R11*4) + MOVL R14, 524288(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeBetterBlockAsm4MB: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeBetterBlockAsm4MB - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x2f, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x08, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x2f, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x08, R11 + IMULQ SI, R11 + SHRQ $0x2f, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeBetterBlockAsm4MB emit_remainder_encodeBetterBlockAsm4MB: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 4(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 4(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBetterBlockAsm4MB - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBetterBlockAsm4MB: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBetterBlockAsm4MB - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBetterBlockAsm4MB @@ -7968,33 +7975,33 @@ emit_remainder_ok_encodeBetterBlockAsm4MB: JB three_bytes_emit_remainder_encodeBetterBlockAsm4MB MOVL DX, BX SHRL $0x10, BX - MOVB $0xf8, (AX) - MOVW DX, 1(AX) - MOVB BL, 3(AX) - ADDQ $0x04, AX + MOVB $0xf8, (CX) + MOVW DX, 1(CX) + MOVB BL, 3(CX) + ADDQ $0x04, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm4MB three_bytes_emit_remainder_encodeBetterBlockAsm4MB: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm4MB two_bytes_emit_remainder_encodeBetterBlockAsm4MB: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBetterBlockAsm4MB JMP memmove_long_emit_remainder_encodeBetterBlockAsm4MB one_byte_emit_remainder_encodeBetterBlockAsm4MB: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBetterBlockAsm4MB: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -8010,73 +8017,73 @@ memmove_emit_remainder_encodeBetterBlockAsm4MB: JMP emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm4MB emit_lit_memmove_emit_remainder_encodeBetterBlockAsm4MB_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBetterBlockAsm4MB: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBetterBlockAsm4MB memmove_long_emit_remainder_encodeBetterBlockAsm4MB: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm4MBlarge_big_loop_back: MOVOU (SI), X4 @@ -8090,756 +8097,757 @@ emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm4MBlarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm4MBlarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm4MBlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBetterBlockAsm4MB: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBetterBlockAsm12B(dst []byte, src []byte) int +// func encodeBetterBlockAsm12B(dst []byte, src []byte, tmp *[81920]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBetterBlockAsm12B(SB), $81944-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000280, CX - LEAQ 24(SP), DX +TEXT ·encodeBetterBlockAsm12B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000280, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBetterBlockAsm12B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBetterBlockAsm12B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -6(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -6(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeBetterBlockAsm12B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x06, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x06, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBetterBlockAsm12B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x34, R10 - MOVL 24(SP)(R9*4), BX - MOVL 65560(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 65560(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x34, R11 + MOVL (AX)(R10*4), SI + MOVL 65536(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 65536(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeBetterBlockAsm12B - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeBetterBlockAsm12B - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeBetterBlockAsm12B no_short_found_encodeBetterBlockAsm12B: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeBetterBlockAsm12B - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeBetterBlockAsm12B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBetterBlockAsm12B candidateS_match_encodeBetterBlockAsm12B: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBetterBlockAsm12B - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeBetterBlockAsm12B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBetterBlockAsm12B match_extend_back_loop_encodeBetterBlockAsm12B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBetterBlockAsm12B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBetterBlockAsm12B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBetterBlockAsm12B JMP match_extend_back_loop_encodeBetterBlockAsm12B match_extend_back_end_encodeBetterBlockAsm12B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBetterBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBetterBlockAsm12B: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeBetterBlockAsm12B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeBetterBlockAsm12B - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm12B - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeBetterBlockAsm12B - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeBetterBlockAsm12B matchlen_bsf_16match_nolit_encodeBetterBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeBetterBlockAsm12B matchlen_match8_match_nolit_encodeBetterBlockAsm12B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeBetterBlockAsm12B - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm12B - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeBetterBlockAsm12B matchlen_bsf_8_match_nolit_encodeBetterBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeBetterBlockAsm12B matchlen_match4_match_nolit_encodeBetterBlockAsm12B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeBetterBlockAsm12B - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeBetterBlockAsm12B - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeBetterBlockAsm12B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeBetterBlockAsm12B JB match_nolit_end_encodeBetterBlockAsm12B - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeBetterBlockAsm12B - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeBetterBlockAsm12B matchlen_match1_match_nolit_encodeBetterBlockAsm12B: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeBetterBlockAsm12B - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeBetterBlockAsm12B: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - CMPL 16(SP), DI + CMPL 16(SP), R8 JEQ match_is_repeat_encodeBetterBlockAsm12B - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeBetterBlockAsm12B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeBetterBlockAsm12B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeBetterBlockAsm12B JB three_bytes_match_emit_encodeBetterBlockAsm12B three_bytes_match_emit_encodeBetterBlockAsm12B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBetterBlockAsm12B two_bytes_match_emit_encodeBetterBlockAsm12B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeBetterBlockAsm12B JMP memmove_long_match_emit_encodeBetterBlockAsm12B one_byte_match_emit_encodeBetterBlockAsm12B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBetterBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_encodeBetterBlockAsm12B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBetterBlockAsm12B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeBetterBlockAsm12B memmove_long_match_emit_encodeBetterBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeBetterBlockAsm12Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBetterBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeBetterBlockAsm12B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeBetterBlockAsm12B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE long_offset_short_match_nolit_encodeBetterBlockAsm12B - MOVL $0x00000001, BX - LEAL 16(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX - SUBL $0x08, R11 + MOVL $0x00000001, SI + LEAL 16(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R12 // emitRepeat - LEAL -4(R11), R11 + LEAL -4(R12), R12 JMP cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_three_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_two_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short_2b: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B long_offset_short_match_nolit_encodeBetterBlockAsm12B: - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm12B_emit_copy_short - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm12B_emit_copy_short - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_three_match_nolit_encodeBetterBlockAsm12B_emit_copy_short: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_two_match_nolit_encodeBetterBlockAsm12B_emit_copy_short: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_two_offset_match_nolit_encodeBetterBlockAsm12B_emit_copy_short: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B two_byte_offset_short_match_nolit_encodeBetterBlockAsm12B: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeBetterBlockAsm12B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeBetterBlockAsm12B - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B emit_copy_three_match_nolit_encodeBetterBlockAsm12B: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B match_is_repeat_encodeBetterBlockAsm12B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_repeat_encodeBetterBlockAsm12B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_repeat_encodeBetterBlockAsm12B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_repeat_encodeBetterBlockAsm12B JB three_bytes_match_emit_repeat_encodeBetterBlockAsm12B three_bytes_match_emit_repeat_encodeBetterBlockAsm12B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm12B two_bytes_match_emit_repeat_encodeBetterBlockAsm12B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_repeat_encodeBetterBlockAsm12B JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm12B one_byte_match_emit_repeat_encodeBetterBlockAsm12B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_repeat_encodeBetterBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_33through64 emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm12B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm12B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm12B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_repeat_encodeBetterBlockAsm12B memmove_long_match_emit_repeat_encodeBetterBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm12Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_repeat_encodeBetterBlockAsm12B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_repeat_encodeBetterBlockAsm12B - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm12B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm12B cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm12B: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_repeat_encodeBetterBlockAsm12B - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_three_match_nolit_repeat_encodeBetterBlockAsm12B: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_two_match_nolit_repeat_encodeBetterBlockAsm12B: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm12B repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm12B: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX match_nolit_emitcopy_end_encodeBetterBlockAsm12B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBetterBlockAsm12B - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBetterBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBetterBlockAsm12B: - MOVQ $0x0000cf1bbcdcbf9b, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x32, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x34, R10 - SHLQ $0x10, R11 - IMULQ BX, R11 - SHRQ $0x32, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x34, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 65560(SP)(R10*4) - MOVL R13, 65560(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x0000cf1bbcdcbf9b, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x10, R10 + IMULQ SI, R10 + SHRQ $0x32, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x34, R11 + SHLQ $0x10, R12 + IMULQ SI, R12 + SHRQ $0x32, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x34, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 65536(AX)(R11*4) + MOVL R14, 65536(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeBetterBlockAsm12B: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeBetterBlockAsm12B - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x32, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x10, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x32, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x10, R11 + IMULQ SI, R11 + SHRQ $0x32, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeBetterBlockAsm12B emit_remainder_encodeBetterBlockAsm12B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBetterBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBetterBlockAsm12B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBetterBlockAsm12B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBetterBlockAsm12B @@ -8848,26 +8856,26 @@ emit_remainder_ok_encodeBetterBlockAsm12B: JB three_bytes_emit_remainder_encodeBetterBlockAsm12B three_bytes_emit_remainder_encodeBetterBlockAsm12B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm12B two_bytes_emit_remainder_encodeBetterBlockAsm12B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBetterBlockAsm12B JMP memmove_long_emit_remainder_encodeBetterBlockAsm12B one_byte_emit_remainder_encodeBetterBlockAsm12B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBetterBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -8883,73 +8891,73 @@ memmove_emit_remainder_encodeBetterBlockAsm12B: JMP emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm12B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBetterBlockAsm12B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBetterBlockAsm12B memmove_long_emit_remainder_encodeBetterBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_big_loop_back: MOVOU (SI), X4 @@ -8963,756 +8971,757 @@ emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 - JAE emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + JAE emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm12Blarge_forward_sse_loop_32 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBetterBlockAsm12B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBetterBlockAsm10B(dst []byte, src []byte) int +// func encodeBetterBlockAsm10B(dst []byte, src []byte, tmp *[20480]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBetterBlockAsm10B(SB), $20504-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x000000a0, CX - LEAQ 24(SP), DX +TEXT ·encodeBetterBlockAsm10B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x000000a0, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBetterBlockAsm10B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBetterBlockAsm10B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -6(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -6(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeBetterBlockAsm10B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBetterBlockAsm10B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x36, R10 - MOVL 24(SP)(R9*4), BX - MOVL 16408(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 16408(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x34, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x36, R11 + MOVL (AX)(R10*4), SI + MOVL 16384(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 16384(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeBetterBlockAsm10B - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeBetterBlockAsm10B - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeBetterBlockAsm10B no_short_found_encodeBetterBlockAsm10B: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeBetterBlockAsm10B - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeBetterBlockAsm10B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBetterBlockAsm10B candidateS_match_encodeBetterBlockAsm10B: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x34, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBetterBlockAsm10B - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeBetterBlockAsm10B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBetterBlockAsm10B match_extend_back_loop_encodeBetterBlockAsm10B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBetterBlockAsm10B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBetterBlockAsm10B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBetterBlockAsm10B JMP match_extend_back_loop_encodeBetterBlockAsm10B match_extend_back_end_encodeBetterBlockAsm10B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBetterBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBetterBlockAsm10B: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeBetterBlockAsm10B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeBetterBlockAsm10B - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm10B - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeBetterBlockAsm10B - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeBetterBlockAsm10B matchlen_bsf_16match_nolit_encodeBetterBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeBetterBlockAsm10B matchlen_match8_match_nolit_encodeBetterBlockAsm10B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeBetterBlockAsm10B - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm10B - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeBetterBlockAsm10B matchlen_bsf_8_match_nolit_encodeBetterBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeBetterBlockAsm10B matchlen_match4_match_nolit_encodeBetterBlockAsm10B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeBetterBlockAsm10B - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeBetterBlockAsm10B - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeBetterBlockAsm10B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeBetterBlockAsm10B JB match_nolit_end_encodeBetterBlockAsm10B - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeBetterBlockAsm10B - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeBetterBlockAsm10B matchlen_match1_match_nolit_encodeBetterBlockAsm10B: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeBetterBlockAsm10B - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeBetterBlockAsm10B: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - CMPL 16(SP), DI + CMPL 16(SP), R8 JEQ match_is_repeat_encodeBetterBlockAsm10B - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeBetterBlockAsm10B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeBetterBlockAsm10B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeBetterBlockAsm10B JB three_bytes_match_emit_encodeBetterBlockAsm10B three_bytes_match_emit_encodeBetterBlockAsm10B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBetterBlockAsm10B two_bytes_match_emit_encodeBetterBlockAsm10B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeBetterBlockAsm10B JMP memmove_long_match_emit_encodeBetterBlockAsm10B one_byte_match_emit_encodeBetterBlockAsm10B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBetterBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_encodeBetterBlockAsm10B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBetterBlockAsm10B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeBetterBlockAsm10B memmove_long_match_emit_encodeBetterBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeBetterBlockAsm10Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBetterBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeBetterBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeBetterBlockAsm10B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeBetterBlockAsm10B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE long_offset_short_match_nolit_encodeBetterBlockAsm10B - MOVL $0x00000001, BX - LEAL 16(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX - SUBL $0x08, R11 + MOVL $0x00000001, SI + LEAL 16(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R12 // emitRepeat - LEAL -4(R11), R11 + LEAL -4(R12), R12 JMP cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_three_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_two_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short_2b: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B long_offset_short_match_nolit_encodeBetterBlockAsm10B: - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm10B_emit_copy_short - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm10B_emit_copy_short - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_three_match_nolit_encodeBetterBlockAsm10B_emit_copy_short: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_two_match_nolit_encodeBetterBlockAsm10B_emit_copy_short: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_two_offset_match_nolit_encodeBetterBlockAsm10B_emit_copy_short: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B two_byte_offset_short_match_nolit_encodeBetterBlockAsm10B: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeBetterBlockAsm10B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeBetterBlockAsm10B - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B emit_copy_three_match_nolit_encodeBetterBlockAsm10B: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B match_is_repeat_encodeBetterBlockAsm10B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_repeat_encodeBetterBlockAsm10B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_repeat_encodeBetterBlockAsm10B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_repeat_encodeBetterBlockAsm10B JB three_bytes_match_emit_repeat_encodeBetterBlockAsm10B three_bytes_match_emit_repeat_encodeBetterBlockAsm10B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm10B two_bytes_match_emit_repeat_encodeBetterBlockAsm10B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_repeat_encodeBetterBlockAsm10B JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm10B one_byte_match_emit_repeat_encodeBetterBlockAsm10B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_repeat_encodeBetterBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_33through64 emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm10B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm10B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm10B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_repeat_encodeBetterBlockAsm10B memmove_long_match_emit_repeat_encodeBetterBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm10Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_repeat_encodeBetterBlockAsm10B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_repeat_encodeBetterBlockAsm10B - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm10B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JB repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm10B cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm10B: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_repeat_encodeBetterBlockAsm10B - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_three_match_nolit_repeat_encodeBetterBlockAsm10B: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_two_match_nolit_repeat_encodeBetterBlockAsm10B: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm10B repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm10B: - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX match_nolit_emitcopy_end_encodeBetterBlockAsm10B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBetterBlockAsm10B - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBetterBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBetterBlockAsm10B: - MOVQ $0x0000cf1bbcdcbf9b, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x34, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x36, R10 - SHLQ $0x10, R11 - IMULQ BX, R11 - SHRQ $0x34, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x36, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 16408(SP)(R10*4) - MOVL R13, 16408(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x0000cf1bbcdcbf9b, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x10, R10 + IMULQ SI, R10 + SHRQ $0x34, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x36, R11 + SHLQ $0x10, R12 + IMULQ SI, R12 + SHRQ $0x34, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x36, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 16384(AX)(R11*4) + MOVL R14, 16384(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeBetterBlockAsm10B: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeBetterBlockAsm10B - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x34, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x10, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x34, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x10, R11 + IMULQ SI, R11 + SHRQ $0x34, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeBetterBlockAsm10B emit_remainder_encodeBetterBlockAsm10B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBetterBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBetterBlockAsm10B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBetterBlockAsm10B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBetterBlockAsm10B @@ -9721,26 +9730,26 @@ emit_remainder_ok_encodeBetterBlockAsm10B: JB three_bytes_emit_remainder_encodeBetterBlockAsm10B three_bytes_emit_remainder_encodeBetterBlockAsm10B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm10B two_bytes_emit_remainder_encodeBetterBlockAsm10B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBetterBlockAsm10B JMP memmove_long_emit_remainder_encodeBetterBlockAsm10B one_byte_emit_remainder_encodeBetterBlockAsm10B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBetterBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -9756,73 +9765,73 @@ memmove_emit_remainder_encodeBetterBlockAsm10B: JMP emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm10B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBetterBlockAsm10B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBetterBlockAsm10B memmove_long_emit_remainder_encodeBetterBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm10Blarge_big_loop_back: MOVOU (SI), X4 @@ -9836,742 +9845,743 @@ emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm10Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBetterBlockAsm10B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeBetterBlockAsm8B(dst []byte, src []byte) int +// func encodeBetterBlockAsm8B(dst []byte, src []byte, tmp *[5120]byte) int // Requires: BMI, SSE2 -TEXT ·encodeBetterBlockAsm8B(SB), $5144-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000028, CX - LEAQ 24(SP), DX +TEXT ·encodeBetterBlockAsm8B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000028, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeBetterBlockAsm8B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeBetterBlockAsm8B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -6(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -6(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeBetterBlockAsm8B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x04, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x04, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeBetterBlockAsm8B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x38, R10 - MOVL 24(SP)(R9*4), BX - MOVL 4120(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 4120(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x36, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x38, R11 + MOVL (AX)(R10*4), SI + MOVL 4096(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 4096(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeBetterBlockAsm8B - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeBetterBlockAsm8B - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeBetterBlockAsm8B no_short_found_encodeBetterBlockAsm8B: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeBetterBlockAsm8B - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeBetterBlockAsm8B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeBetterBlockAsm8B candidateS_match_encodeBetterBlockAsm8B: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x36, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeBetterBlockAsm8B - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeBetterBlockAsm8B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeBetterBlockAsm8B match_extend_back_loop_encodeBetterBlockAsm8B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeBetterBlockAsm8B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeBetterBlockAsm8B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeBetterBlockAsm8B JMP match_extend_back_loop_encodeBetterBlockAsm8B match_extend_back_end_encodeBetterBlockAsm8B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeBetterBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeBetterBlockAsm8B: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeBetterBlockAsm8B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeBetterBlockAsm8B - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm8B - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeBetterBlockAsm8B - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeBetterBlockAsm8B matchlen_bsf_16match_nolit_encodeBetterBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeBetterBlockAsm8B matchlen_match8_match_nolit_encodeBetterBlockAsm8B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeBetterBlockAsm8B - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeBetterBlockAsm8B - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeBetterBlockAsm8B matchlen_bsf_8_match_nolit_encodeBetterBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeBetterBlockAsm8B matchlen_match4_match_nolit_encodeBetterBlockAsm8B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeBetterBlockAsm8B - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeBetterBlockAsm8B - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeBetterBlockAsm8B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeBetterBlockAsm8B JB match_nolit_end_encodeBetterBlockAsm8B - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeBetterBlockAsm8B - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeBetterBlockAsm8B matchlen_match1_match_nolit_encodeBetterBlockAsm8B: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeBetterBlockAsm8B - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeBetterBlockAsm8B: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - CMPL 16(SP), DI + CMPL 16(SP), R8 JEQ match_is_repeat_encodeBetterBlockAsm8B - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeBetterBlockAsm8B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeBetterBlockAsm8B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeBetterBlockAsm8B JB three_bytes_match_emit_encodeBetterBlockAsm8B three_bytes_match_emit_encodeBetterBlockAsm8B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeBetterBlockAsm8B two_bytes_match_emit_encodeBetterBlockAsm8B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeBetterBlockAsm8B JMP memmove_long_match_emit_encodeBetterBlockAsm8B one_byte_match_emit_encodeBetterBlockAsm8B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeBetterBlockAsm8B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x04 + CMPQ R9, $0x04 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_4 - CMPQ R8, $0x08 + CMPQ R9, $0x08 JB emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_4through7 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_4: - MOVL (R9), R10 - MOVL R10, (AX) + MOVL (R10), R11 + MOVL R11, (CX) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_4through7: - MOVL (R9), R10 - MOVL -4(R9)(R8*1), R9 - MOVL R10, (AX) - MOVL R9, -4(AX)(R8*1) + MOVL (R10), R11 + MOVL -4(R10)(R9*1), R10 + MOVL R11, (CX) + MOVL R10, -4(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_encodeBetterBlockAsm8B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeBetterBlockAsm8B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeBetterBlockAsm8B memmove_long_match_emit_encodeBetterBlockAsm8B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeBetterBlockAsm8Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeBetterBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeBetterBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeBetterBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeBetterBlockAsm8B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeBetterBlockAsm8B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE long_offset_short_match_nolit_encodeBetterBlockAsm8B - MOVL $0x00000001, BX - LEAL 16(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX - SUBL $0x08, R11 + MOVL $0x00000001, SI + LEAL 16(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX + SUBL $0x08, R12 // emitRepeat - LEAL -4(R11), R11 + LEAL -4(R12), R12 JMP cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B repeat_three_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B repeat_two_match_nolit_encodeBetterBlockAsm8B_emit_copy_short_2b: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B long_offset_short_match_nolit_encodeBetterBlockAsm8B: - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_encodeBetterBlockAsm8B_emit_copy_short - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm8B_emit_copy_short cant_repeat_two_offset_match_nolit_encodeBetterBlockAsm8B_emit_copy_short: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_encodeBetterBlockAsm8B_emit_copy_short - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B repeat_three_match_nolit_encodeBetterBlockAsm8B_emit_copy_short: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B -repeat_two_match_nolit_encodeBetterBlockAsm8B_emit_copy_short: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX - JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX +repeat_two_match_nolit_encodeBetterBlockAsm8B_emit_copy_short: + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX + JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B two_byte_offset_short_match_nolit_encodeBetterBlockAsm8B: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeBetterBlockAsm8B - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B emit_copy_three_match_nolit_encodeBetterBlockAsm8B: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B match_is_repeat_encodeBetterBlockAsm8B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_repeat_encodeBetterBlockAsm8B - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_match_emit_repeat_encodeBetterBlockAsm8B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_repeat_encodeBetterBlockAsm8B JB three_bytes_match_emit_repeat_encodeBetterBlockAsm8B three_bytes_match_emit_repeat_encodeBetterBlockAsm8B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm8B two_bytes_match_emit_repeat_encodeBetterBlockAsm8B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_repeat_encodeBetterBlockAsm8B JMP memmove_long_match_emit_repeat_encodeBetterBlockAsm8B one_byte_match_emit_repeat_encodeBetterBlockAsm8B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_repeat_encodeBetterBlockAsm8B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveShort - CMPQ DI, $0x04 + CMPQ R8, $0x04 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_4 - CMPQ DI, $0x08 + CMPQ R8, $0x08 JB emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_4through7 - CMPQ DI, $0x10 + CMPQ R8, $0x10 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_8through16 - CMPQ DI, $0x20 + CMPQ R8, $0x20 JBE emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_33through64 emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_4: - MOVL (R8), R9 - MOVL R9, (AX) + MOVL (R9), R10 + MOVL R10, (CX) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_4through7: - MOVL (R8), R9 - MOVL -4(R8)(DI*1), R8 - MOVL R9, (AX) - MOVL R8, -4(AX)(DI*1) + MOVL (R9), R10 + MOVL -4(R9)(R8*1), R9 + MOVL R10, (CX) + MOVL R9, -4(CX)(R8*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_8through16: - MOVQ (R8), R9 - MOVQ -8(R8)(DI*1), R8 - MOVQ R9, (AX) - MOVQ R8, -8(AX)(DI*1) + MOVQ (R9), R10 + MOVQ -8(R9)(R8*1), R9 + MOVQ R10, (CX) + MOVQ R9, -8(CX)(R8*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_17through32: - MOVOU (R8), X0 - MOVOU -16(R8)(DI*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU -16(R9)(R8*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R8*1) JMP memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm8B emit_lit_memmove_match_emit_repeat_encodeBetterBlockAsm8B_memmove_move_33through64: - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) memmove_end_copy_match_emit_repeat_encodeBetterBlockAsm8B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_repeat_encodeBetterBlockAsm8B memmove_long_match_emit_repeat_encodeBetterBlockAsm8B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveLong - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVQ DI, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R12 - SUBQ R9, R12 - DECQ R10 + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVQ R8, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R13 + SUBQ R10, R13 + DECQ R11 JA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(R8)(R12*1), R9 - LEAQ -32(AX)(R12*1), R13 + LEAQ -32(R9)(R13*1), R10 + LEAQ -32(CX)(R13*1), R14 emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm8Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R13) - MOVOA X5, 16(R13) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R14) + MOVOA X5, 16(R14) + ADDQ $0x20, R14 + ADDQ $0x20, R10 ADDQ $0x20, R13 - ADDQ $0x20, R9 - ADDQ $0x20, R12 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(R8)(R12*1), X4 - MOVOU -16(R8)(R12*1), X5 - MOVOA X4, -32(AX)(R12*1) - MOVOA X5, -16(AX)(R12*1) - ADDQ $0x20, R12 - CMPQ DI, R12 + MOVOU -32(R9)(R13*1), X4 + MOVOU -16(R9)(R13*1), X5 + MOVOA X4, -32(CX)(R13*1) + MOVOA X5, -16(CX)(R13*1) + ADDQ $0x20, R13 + CMPQ R8, R13 JAE emit_lit_memmove_long_match_emit_repeat_encodeBetterBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) + MOVQ SI, CX emit_literal_done_match_emit_repeat_encodeBetterBlockAsm8B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitRepeat - MOVL R11, BX - LEAL -4(R11), R11 - CMPL BX, $0x08 + MOVL R12, SI + LEAL -4(R12), R12 + CMPL SI, $0x08 JBE repeat_two_match_nolit_repeat_encodeBetterBlockAsm8B - CMPL BX, $0x0c + CMPL SI, $0x0c JAE cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm8B cant_repeat_two_offset_match_nolit_repeat_encodeBetterBlockAsm8B: - CMPL R11, $0x00000104 + CMPL R12, $0x00000104 JB repeat_three_match_nolit_repeat_encodeBetterBlockAsm8B - LEAL -256(R11), R11 - MOVW $0x0019, (AX) - MOVW R11, 2(AX) - ADDQ $0x04, AX + LEAL -256(R12), R12 + MOVW $0x0019, (CX) + MOVW R12, 2(CX) + ADDQ $0x04, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B repeat_three_match_nolit_repeat_encodeBetterBlockAsm8B: - LEAL -4(R11), R11 - MOVW $0x0015, (AX) - MOVB R11, 2(AX) - ADDQ $0x03, AX + LEAL -4(R12), R12 + MOVW $0x0015, (CX) + MOVB R12, 2(CX) + ADDQ $0x03, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B repeat_two_match_nolit_repeat_encodeBetterBlockAsm8B: - SHLL $0x02, R11 - ORL $0x01, R11 - MOVW R11, (AX) - ADDQ $0x02, AX + SHLL $0x02, R12 + ORL $0x01, R12 + MOVW R12, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeBetterBlockAsm8B - XORQ BX, BX - LEAL 1(BX)(R11*4), R11 - MOVB DI, 1(AX) - SARL $0x08, DI - SHLL $0x05, DI - ORL DI, R11 - MOVB R11, (AX) - ADDQ $0x02, AX + XORQ SI, SI + LEAL 1(SI)(R12*4), R12 + MOVB R8, 1(CX) + SARL $0x08, R8 + SHLL $0x05, R8 + ORL R8, R12 + MOVB R12, (CX) + ADDQ $0x02, CX match_nolit_emitcopy_end_encodeBetterBlockAsm8B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeBetterBlockAsm8B - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeBetterBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeBetterBlockAsm8B: - MOVQ $0x0000cf1bbcdcbf9b, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x36, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x38, R10 - SHLQ $0x10, R11 - IMULQ BX, R11 - SHRQ $0x36, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x38, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 4120(SP)(R10*4) - MOVL R13, 4120(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x0000cf1bbcdcbf9b, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x10, R10 + IMULQ SI, R10 + SHRQ $0x36, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x38, R11 + SHLQ $0x10, R12 + IMULQ SI, R12 + SHRQ $0x36, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x38, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 4096(AX)(R11*4) + MOVL R14, 4096(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeBetterBlockAsm8B: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeBetterBlockAsm8B - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x36, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x10, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x36, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x10, R11 + IMULQ SI, R11 + SHRQ $0x36, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeBetterBlockAsm8B emit_remainder_encodeBetterBlockAsm8B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeBetterBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeBetterBlockAsm8B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeBetterBlockAsm8B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeBetterBlockAsm8B @@ -10580,26 +10590,26 @@ emit_remainder_ok_encodeBetterBlockAsm8B: JB three_bytes_emit_remainder_encodeBetterBlockAsm8B three_bytes_emit_remainder_encodeBetterBlockAsm8B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeBetterBlockAsm8B two_bytes_emit_remainder_encodeBetterBlockAsm8B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeBetterBlockAsm8B JMP memmove_long_emit_remainder_encodeBetterBlockAsm8B one_byte_emit_remainder_encodeBetterBlockAsm8B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeBetterBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -10615,73 +10625,73 @@ memmove_emit_remainder_encodeBetterBlockAsm8B: JMP emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeBetterBlockAsm8B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeBetterBlockAsm8B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeBetterBlockAsm8B memmove_long_emit_remainder_encodeBetterBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm8Blarge_big_loop_back: MOVOU (SI), X4 @@ -10695,798 +10705,799 @@ emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm8Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeBetterBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeBetterBlockAsm8B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBlockAsm(dst []byte, src []byte) int +// func encodeSnappyBlockAsm(dst []byte, src []byte, tmp *[65536]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBlockAsm(SB), $65560-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000200, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBlockAsm(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBlockAsm: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBlockAsm MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBlockAsm: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x06, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x06, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x10, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x10, R11 + IMULQ R9, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeSnappyBlockAsm - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_encodeSnappyBlockAsm repeat_extend_back_loop_encodeSnappyBlockAsm: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_encodeSnappyBlockAsm - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeSnappyBlockAsm - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_encodeSnappyBlockAsm repeat_extend_back_end_encodeSnappyBlockAsm: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 5(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 5(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeSnappyBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeSnappyBlockAsm: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeSnappyBlockAsm - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeSnappyBlockAsm - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeSnappyBlockAsm - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_repeat_emit_encodeSnappyBlockAsm - CMPL BX, $0x01000000 + CMPL SI, $0x01000000 JB four_bytes_repeat_emit_encodeSnappyBlockAsm - MOVB $0xfc, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm four_bytes_repeat_emit_encodeSnappyBlockAsm: - MOVL BX, R9 - SHRL $0x10, R9 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R9, 3(AX) - ADDQ $0x04, AX + MOVL SI, R10 + SHRL $0x10, R10 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R10, 3(CX) + ADDQ $0x04, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm three_bytes_repeat_emit_encodeSnappyBlockAsm: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm two_bytes_repeat_emit_encodeSnappyBlockAsm: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeSnappyBlockAsm JMP memmove_long_repeat_emit_encodeSnappyBlockAsm one_byte_repeat_emit_encodeSnappyBlockAsm: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeSnappyBlockAsm: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveShort - CMPQ DI, $0x08 + CMPQ R8, $0x08 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_8 - CMPQ DI, $0x10 + CMPQ R8, $0x10 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_8through16 - CMPQ DI, $0x20 + CMPQ R8, $0x20 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_8: - MOVQ (R8), R9 - MOVQ R9, (AX) + MOVQ (R9), R10 + MOVQ R10, (CX) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_8through16: - MOVQ (R8), R9 - MOVQ -8(R8)(DI*1), R8 - MOVQ R9, (AX) - MOVQ R8, -8(AX)(DI*1) + MOVQ (R9), R10 + MOVQ -8(R9)(R8*1), R9 + MOVQ R10, (CX) + MOVQ R9, -8(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_17through32: - MOVOU (R8), X0 - MOVOU -16(R8)(DI*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU -16(R9)(R8*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm_memmove_move_33through64: - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) memmove_end_copy_repeat_emit_encodeSnappyBlockAsm: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeSnappyBlockAsm memmove_long_repeat_emit_encodeSnappyBlockAsm: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveLong - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVQ DI, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVQ R8, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(R8)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(R9)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsmlarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsmlarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(R8)(R11*1), X4 - MOVOU -16(R8)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ DI, R11 + MOVOU -32(R9)(R12*1), X4 + MOVOU -16(R9)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R8, R12 JAE emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeSnappyBlockAsm: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_encodeSnappyBlockAsm - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm matchlen_match8_repeat_extend_encodeSnappyBlockAsm: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_encodeSnappyBlockAsm - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_encodeSnappyBlockAsm matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm matchlen_match4_repeat_extend_encodeSnappyBlockAsm: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_encodeSnappyBlockAsm - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_encodeSnappyBlockAsm - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_encodeSnappyBlockAsm: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_encodeSnappyBlockAsm JB repeat_extend_forward_end_encodeSnappyBlockAsm - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_encodeSnappyBlockAsm - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_encodeSnappyBlockAsm matchlen_match1_repeat_extend_encodeSnappyBlockAsm: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_encodeSnappyBlockAsm - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_encodeSnappyBlockAsm: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy - CMPL SI, $0x00010000 + CMPL DI, $0x00010000 JB two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm four_bytes_loop_back_repeat_as_copy_encodeSnappyBlockAsm: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE four_bytes_remain_repeat_as_copy_encodeSnappyBlockAsm - MOVB $0xff, (AX) - MOVL SI, 1(AX) - LEAL -64(BX), BX - ADDQ $0x05, AX - CMPL BX, $0x04 + MOVB $0xff, (CX) + MOVL DI, 1(CX) + LEAL -64(SI), SI + ADDQ $0x05, CX + CMPL SI, $0x04 JB four_bytes_remain_repeat_as_copy_encodeSnappyBlockAsm JMP four_bytes_loop_back_repeat_as_copy_encodeSnappyBlockAsm four_bytes_remain_repeat_as_copy_encodeSnappyBlockAsm: - TESTL BX, BX + TESTL SI, SI JZ repeat_end_emit_encodeSnappyBlockAsm - XORL DI, DI - LEAL -1(DI)(BX*4), BX - MOVB BL, (AX) - MOVL SI, 1(AX) - ADDQ $0x05, AX + XORL R8, R8 + LEAL -1(R8)(SI*4), SI + MOVB SI, (CX) + MOVL DI, 1(CX) + ADDQ $0x05, CX JMP repeat_end_emit_encodeSnappyBlockAsm two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeSnappyBlockAsm emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeSnappyBlockAsm: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeSnappyBlockAsm no_repeat_found_encodeSnappyBlockAsm: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBlockAsm - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeSnappyBlockAsm - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeSnappyBlockAsm - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBlockAsm candidate3_match_encodeSnappyBlockAsm: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeSnappyBlockAsm candidate2_match_encodeSnappyBlockAsm: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeSnappyBlockAsm: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBlockAsm match_extend_back_loop_encodeSnappyBlockAsm: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBlockAsm - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBlockAsm - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBlockAsm JMP match_extend_back_loop_encodeSnappyBlockAsm match_extend_back_end_encodeSnappyBlockAsm: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 5(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 5(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBlockAsm: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeSnappyBlockAsm - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeSnappyBlockAsm - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeSnappyBlockAsm - CMPL DI, $0x00010000 + CMPL R8, $0x00010000 JB three_bytes_match_emit_encodeSnappyBlockAsm - CMPL DI, $0x01000000 + CMPL R8, $0x01000000 JB four_bytes_match_emit_encodeSnappyBlockAsm - MOVB $0xfc, (AX) - MOVL DI, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL R8, 1(CX) + ADDQ $0x05, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm four_bytes_match_emit_encodeSnappyBlockAsm: - MOVL DI, R9 - SHRL $0x10, R9 - MOVB $0xf8, (AX) - MOVW DI, 1(AX) - MOVB R9, 3(AX) - ADDQ $0x04, AX + MOVL R8, R10 + SHRL $0x10, R10 + MOVB $0xf8, (CX) + MOVW R8, 1(CX) + MOVB R10, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm three_bytes_match_emit_encodeSnappyBlockAsm: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm two_bytes_match_emit_encodeSnappyBlockAsm: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeSnappyBlockAsm JMP memmove_long_match_emit_encodeSnappyBlockAsm one_byte_match_emit_encodeSnappyBlockAsm: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBlockAsm: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm emit_lit_memmove_match_emit_encodeSnappyBlockAsm_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBlockAsm: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeSnappyBlockAsm memmove_long_match_emit_encodeSnappyBlockAsm: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeSnappyBlockAsmlarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsmlarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeSnappyBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeSnappyBlockAsm: match_nolit_loop_encodeSnappyBlockAsm: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBlockAsm - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeSnappyBlockAsm - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm matchlen_bsf_16match_nolit_encodeSnappyBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm matchlen_match8_match_nolit_encodeSnappyBlockAsm: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBlockAsm - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeSnappyBlockAsm matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm matchlen_match4_match_nolit_encodeSnappyBlockAsm: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBlockAsm - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeSnappyBlockAsm - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeSnappyBlockAsm: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBlockAsm JB match_nolit_end_encodeSnappyBlockAsm - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeSnappyBlockAsm - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeSnappyBlockAsm matchlen_match1_match_nolit_encodeSnappyBlockAsm: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeSnappyBlockAsm - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeSnappyBlockAsm: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB two_byte_offset_match_nolit_encodeSnappyBlockAsm four_bytes_loop_back_match_nolit_encodeSnappyBlockAsm: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE four_bytes_remain_match_nolit_encodeSnappyBlockAsm - MOVB $0xff, (AX) - MOVL BX, 1(AX) - LEAL -64(R9), R9 - ADDQ $0x05, AX - CMPL R9, $0x04 + MOVB $0xff, (CX) + MOVL SI, 1(CX) + LEAL -64(R10), R10 + ADDQ $0x05, CX + CMPL R10, $0x04 JB four_bytes_remain_match_nolit_encodeSnappyBlockAsm JMP four_bytes_loop_back_match_nolit_encodeSnappyBlockAsm four_bytes_remain_match_nolit_encodeSnappyBlockAsm: - TESTL R9, R9 + TESTL R10, R10 JZ match_nolit_emitcopy_end_encodeSnappyBlockAsm - XORL SI, SI - LEAL -1(SI)(R9*4), R9 - MOVB R9, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + XORL DI, DI + LEAL -1(DI)(R10*4), R10 + MOVB R10, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeSnappyBlockAsm two_byte_offset_match_nolit_encodeSnappyBlockAsm: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBlockAsm - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBlockAsm two_byte_offset_short_match_nolit_encodeSnappyBlockAsm: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBlockAsm emit_copy_three_match_nolit_encodeSnappyBlockAsm: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBlockAsm: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBlockAsm: - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x10, DI - IMULQ R8, DI - SHRQ $0x32, DI - SHLQ $0x10, BX - IMULQ R8, BX - SHRQ $0x32, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x10, R8 + IMULQ R9, R8 + SHRQ $0x32, R8 + SHLQ $0x10, SI + IMULQ R9, SI + SHRQ $0x32, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeSnappyBlockAsm - INCL CX + INCL DX JMP search_loop_encodeSnappyBlockAsm emit_remainder_encodeSnappyBlockAsm: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 5(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 5(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBlockAsm: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBlockAsm - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBlockAsm @@ -11496,41 +11507,41 @@ emit_remainder_ok_encodeSnappyBlockAsm: JB three_bytes_emit_remainder_encodeSnappyBlockAsm CMPL DX, $0x01000000 JB four_bytes_emit_remainder_encodeSnappyBlockAsm - MOVB $0xfc, (AX) - MOVL DX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL DX, 1(CX) + ADDQ $0x05, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm four_bytes_emit_remainder_encodeSnappyBlockAsm: MOVL DX, BX SHRL $0x10, BX - MOVB $0xf8, (AX) - MOVW DX, 1(AX) - MOVB BL, 3(AX) - ADDQ $0x04, AX + MOVB $0xf8, (CX) + MOVW DX, 1(CX) + MOVB BL, 3(CX) + ADDQ $0x04, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm three_bytes_emit_remainder_encodeSnappyBlockAsm: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm two_bytes_emit_remainder_encodeSnappyBlockAsm: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBlockAsm JMP memmove_long_emit_remainder_encodeSnappyBlockAsm one_byte_emit_remainder_encodeSnappyBlockAsm: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -11546,73 +11557,73 @@ memmove_emit_remainder_encodeSnappyBlockAsm: JMP emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBlockAsm: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBlockAsm memmove_long_emit_remainder_encodeSnappyBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsmlarge_big_loop_back: MOVOU (SI), X4 @@ -11626,718 +11637,719 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsmlarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsmlarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBlockAsm: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBlockAsm64K(dst []byte, src []byte) int +// func encodeSnappyBlockAsm64K(dst []byte, src []byte, tmp *[65536]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBlockAsm64K(SB), $65560-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000200, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBlockAsm64K(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBlockAsm64K: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBlockAsm64K MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBlockAsm64K: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x06, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x06, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm64K - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x10, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x10, R11 + IMULQ R9, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeSnappyBlockAsm64K - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_encodeSnappyBlockAsm64K repeat_extend_back_loop_encodeSnappyBlockAsm64K: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_encodeSnappyBlockAsm64K - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeSnappyBlockAsm64K - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_encodeSnappyBlockAsm64K repeat_extend_back_end_encodeSnappyBlockAsm64K: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeSnappyBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeSnappyBlockAsm64K: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeSnappyBlockAsm64K - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeSnappyBlockAsm64K - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeSnappyBlockAsm64K JB three_bytes_repeat_emit_encodeSnappyBlockAsm64K three_bytes_repeat_emit_encodeSnappyBlockAsm64K: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm64K two_bytes_repeat_emit_encodeSnappyBlockAsm64K: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeSnappyBlockAsm64K JMP memmove_long_repeat_emit_encodeSnappyBlockAsm64K one_byte_repeat_emit_encodeSnappyBlockAsm64K: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeSnappyBlockAsm64K: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveShort - CMPQ DI, $0x08 + CMPQ R8, $0x08 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_8 - CMPQ DI, $0x10 + CMPQ R8, $0x10 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_8through16 - CMPQ DI, $0x20 + CMPQ R8, $0x20 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_8: - MOVQ (R8), R9 - MOVQ R9, (AX) + MOVQ (R9), R10 + MOVQ R10, (CX) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm64K emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_8through16: - MOVQ (R8), R9 - MOVQ -8(R8)(DI*1), R8 - MOVQ R9, (AX) - MOVQ R8, -8(AX)(DI*1) + MOVQ (R9), R10 + MOVQ -8(R9)(R8*1), R9 + MOVQ R10, (CX) + MOVQ R9, -8(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm64K emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_17through32: - MOVOU (R8), X0 - MOVOU -16(R8)(DI*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU -16(R9)(R8*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm64K emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm64K_memmove_move_33through64: - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) memmove_end_copy_repeat_emit_encodeSnappyBlockAsm64K: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeSnappyBlockAsm64K memmove_long_repeat_emit_encodeSnappyBlockAsm64K: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveLong - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVQ DI, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVQ R8, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32 - LEAQ -32(R8)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(R9)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm64Klarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm64Klarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32: - MOVOU -32(R8)(R11*1), X4 - MOVOU -16(R8)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ DI, R11 + MOVOU -32(R9)(R12*1), X4 + MOVOU -16(R9)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R8, R12 JAE emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeSnappyBlockAsm64K: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm64K: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_encodeSnappyBlockAsm64K - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm64K - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm64K - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm64K matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm64K: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm64K matchlen_match8_repeat_extend_encodeSnappyBlockAsm64K: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_encodeSnappyBlockAsm64K - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm64K - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_encodeSnappyBlockAsm64K matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm64K: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm64K matchlen_match4_repeat_extend_encodeSnappyBlockAsm64K: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_encodeSnappyBlockAsm64K - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_encodeSnappyBlockAsm64K - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_encodeSnappyBlockAsm64K: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_encodeSnappyBlockAsm64K JB repeat_extend_forward_end_encodeSnappyBlockAsm64K - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_encodeSnappyBlockAsm64K - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_encodeSnappyBlockAsm64K matchlen_match1_repeat_extend_encodeSnappyBlockAsm64K: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_encodeSnappyBlockAsm64K - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_encodeSnappyBlockAsm64K: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm64K: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm64K - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm64K two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm64K: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm64K - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm64K - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeSnappyBlockAsm64K emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm64K: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeSnappyBlockAsm64K: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeSnappyBlockAsm64K no_repeat_found_encodeSnappyBlockAsm64K: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBlockAsm64K - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeSnappyBlockAsm64K - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeSnappyBlockAsm64K - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBlockAsm64K candidate3_match_encodeSnappyBlockAsm64K: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeSnappyBlockAsm64K candidate2_match_encodeSnappyBlockAsm64K: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeSnappyBlockAsm64K: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBlockAsm64K match_extend_back_loop_encodeSnappyBlockAsm64K: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBlockAsm64K - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBlockAsm64K - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBlockAsm64K JMP match_extend_back_loop_encodeSnappyBlockAsm64K match_extend_back_end_encodeSnappyBlockAsm64K: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBlockAsm64K: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeSnappyBlockAsm64K - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeSnappyBlockAsm64K - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeSnappyBlockAsm64K JB three_bytes_match_emit_encodeSnappyBlockAsm64K three_bytes_match_emit_encodeSnappyBlockAsm64K: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm64K two_bytes_match_emit_encodeSnappyBlockAsm64K: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeSnappyBlockAsm64K JMP memmove_long_match_emit_encodeSnappyBlockAsm64K one_byte_match_emit_encodeSnappyBlockAsm64K: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBlockAsm64K: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm64K emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm64K emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm64K emit_lit_memmove_match_emit_encodeSnappyBlockAsm64K_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBlockAsm64K: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeSnappyBlockAsm64K memmove_long_match_emit_encodeSnappyBlockAsm64K: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm64Klarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm64Klarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeSnappyBlockAsm64K: match_nolit_loop_encodeSnappyBlockAsm64K: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm64K: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBlockAsm64K - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm64K - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeSnappyBlockAsm64K - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm64K matchlen_bsf_16match_nolit_encodeSnappyBlockAsm64K: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm64K matchlen_match8_match_nolit_encodeSnappyBlockAsm64K: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBlockAsm64K - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm64K - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeSnappyBlockAsm64K matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm64K: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm64K matchlen_match4_match_nolit_encodeSnappyBlockAsm64K: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBlockAsm64K - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeSnappyBlockAsm64K - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeSnappyBlockAsm64K: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBlockAsm64K JB match_nolit_end_encodeSnappyBlockAsm64K - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeSnappyBlockAsm64K - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeSnappyBlockAsm64K matchlen_match1_match_nolit_encodeSnappyBlockAsm64K: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeSnappyBlockAsm64K - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeSnappyBlockAsm64K: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBlockAsm64K: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBlockAsm64K - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBlockAsm64K two_byte_offset_short_match_nolit_encodeSnappyBlockAsm64K: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm64K - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm64K - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBlockAsm64K emit_copy_three_match_nolit_encodeSnappyBlockAsm64K: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBlockAsm64K: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm64K - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBlockAsm64K: - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x10, DI - IMULQ R8, DI - SHRQ $0x32, DI - SHLQ $0x10, BX - IMULQ R8, BX - SHRQ $0x32, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x10, R8 + IMULQ R9, R8 + SHRQ $0x32, R8 + SHLQ $0x10, SI + IMULQ R9, SI + SHRQ $0x32, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeSnappyBlockAsm64K - INCL CX + INCL DX JMP search_loop_encodeSnappyBlockAsm64K emit_remainder_encodeSnappyBlockAsm64K: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBlockAsm64K: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBlockAsm64K - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBlockAsm64K @@ -12346,26 +12358,26 @@ emit_remainder_ok_encodeSnappyBlockAsm64K: JB three_bytes_emit_remainder_encodeSnappyBlockAsm64K three_bytes_emit_remainder_encodeSnappyBlockAsm64K: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm64K two_bytes_emit_remainder_encodeSnappyBlockAsm64K: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBlockAsm64K JMP memmove_long_emit_remainder_encodeSnappyBlockAsm64K one_byte_emit_remainder_encodeSnappyBlockAsm64K: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBlockAsm64K: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -12381,73 +12393,73 @@ memmove_emit_remainder_encodeSnappyBlockAsm64K: JMP emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm64K -emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) +emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_4through7: + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm64K_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBlockAsm64K: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBlockAsm64K memmove_long_emit_remainder_encodeSnappyBlockAsm64K: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm64Klarge_big_loop_back: MOVOU (SI), X4 @@ -12461,718 +12473,719 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm64Klarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm64Klarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm64Klarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBlockAsm64K: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBlockAsm12B(dst []byte, src []byte) int +// func encodeSnappyBlockAsm12B(dst []byte, src []byte, tmp *[16384]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBlockAsm12B(SB), $16408-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000080, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBlockAsm12B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000080, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBlockAsm12B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBlockAsm12B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBlockAsm12B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm12B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x000000cf1bbcdcbb, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x18, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x000000cf1bbcdcbb, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x18, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x34, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x18, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x18, R11 + IMULQ R9, R11 + SHRQ $0x34, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x18, R10 + IMULQ R9, R10 + SHRQ $0x34, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeSnappyBlockAsm12B - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_encodeSnappyBlockAsm12B repeat_extend_back_loop_encodeSnappyBlockAsm12B: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_encodeSnappyBlockAsm12B - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeSnappyBlockAsm12B - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_encodeSnappyBlockAsm12B repeat_extend_back_end_encodeSnappyBlockAsm12B: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeSnappyBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeSnappyBlockAsm12B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeSnappyBlockAsm12B - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeSnappyBlockAsm12B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeSnappyBlockAsm12B JB three_bytes_repeat_emit_encodeSnappyBlockAsm12B three_bytes_repeat_emit_encodeSnappyBlockAsm12B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm12B two_bytes_repeat_emit_encodeSnappyBlockAsm12B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeSnappyBlockAsm12B JMP memmove_long_repeat_emit_encodeSnappyBlockAsm12B one_byte_repeat_emit_encodeSnappyBlockAsm12B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeSnappyBlockAsm12B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveShort - CMPQ DI, $0x08 + CMPQ R8, $0x08 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_8 - CMPQ DI, $0x10 + CMPQ R8, $0x10 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_8through16 - CMPQ DI, $0x20 + CMPQ R8, $0x20 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_8: - MOVQ (R8), R9 - MOVQ R9, (AX) + MOVQ (R9), R10 + MOVQ R10, (CX) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm12B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_8through16: - MOVQ (R8), R9 - MOVQ -8(R8)(DI*1), R8 - MOVQ R9, (AX) - MOVQ R8, -8(AX)(DI*1) + MOVQ (R9), R10 + MOVQ -8(R9)(R8*1), R9 + MOVQ R10, (CX) + MOVQ R9, -8(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm12B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_17through32: - MOVOU (R8), X0 - MOVOU -16(R8)(DI*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU -16(R9)(R8*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm12B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm12B_memmove_move_33through64: - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) memmove_end_copy_repeat_emit_encodeSnappyBlockAsm12B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeSnappyBlockAsm12B memmove_long_repeat_emit_encodeSnappyBlockAsm12B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveLong - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVQ DI, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVQ R8, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(R8)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(R9)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm12Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(R8)(R11*1), X4 - MOVOU -16(R8)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ DI, R11 + MOVOU -32(R9)(R12*1), X4 + MOVOU -16(R9)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R8, R12 JAE emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeSnappyBlockAsm12B: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm12B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_encodeSnappyBlockAsm12B - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm12B - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm12B - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm12B matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm12B matchlen_match8_repeat_extend_encodeSnappyBlockAsm12B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_encodeSnappyBlockAsm12B - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm12B - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_encodeSnappyBlockAsm12B matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm12B matchlen_match4_repeat_extend_encodeSnappyBlockAsm12B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_encodeSnappyBlockAsm12B - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_encodeSnappyBlockAsm12B - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_encodeSnappyBlockAsm12B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_encodeSnappyBlockAsm12B JB repeat_extend_forward_end_encodeSnappyBlockAsm12B - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_encodeSnappyBlockAsm12B - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_encodeSnappyBlockAsm12B matchlen_match1_repeat_extend_encodeSnappyBlockAsm12B: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_encodeSnappyBlockAsm12B - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_encodeSnappyBlockAsm12B: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm12B: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm12B - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm12B two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm12B: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm12B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm12B - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeSnappyBlockAsm12B emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm12B: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeSnappyBlockAsm12B: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeSnappyBlockAsm12B no_repeat_found_encodeSnappyBlockAsm12B: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBlockAsm12B - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeSnappyBlockAsm12B - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeSnappyBlockAsm12B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBlockAsm12B candidate3_match_encodeSnappyBlockAsm12B: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeSnappyBlockAsm12B candidate2_match_encodeSnappyBlockAsm12B: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeSnappyBlockAsm12B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBlockAsm12B match_extend_back_loop_encodeSnappyBlockAsm12B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBlockAsm12B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBlockAsm12B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBlockAsm12B JMP match_extend_back_loop_encodeSnappyBlockAsm12B match_extend_back_end_encodeSnappyBlockAsm12B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBlockAsm12B: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeSnappyBlockAsm12B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeSnappyBlockAsm12B - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeSnappyBlockAsm12B JB three_bytes_match_emit_encodeSnappyBlockAsm12B three_bytes_match_emit_encodeSnappyBlockAsm12B: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm12B two_bytes_match_emit_encodeSnappyBlockAsm12B: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeSnappyBlockAsm12B JMP memmove_long_match_emit_encodeSnappyBlockAsm12B one_byte_match_emit_encodeSnappyBlockAsm12B: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBlockAsm12B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm12B emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm12B emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm12B emit_lit_memmove_match_emit_encodeSnappyBlockAsm12B_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBlockAsm12B: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeSnappyBlockAsm12B memmove_long_match_emit_encodeSnappyBlockAsm12B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm12Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeSnappyBlockAsm12B: match_nolit_loop_encodeSnappyBlockAsm12B: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm12B: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBlockAsm12B - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm12B - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeSnappyBlockAsm12B - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm12B matchlen_bsf_16match_nolit_encodeSnappyBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm12B matchlen_match8_match_nolit_encodeSnappyBlockAsm12B: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBlockAsm12B - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm12B - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeSnappyBlockAsm12B matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm12B matchlen_match4_match_nolit_encodeSnappyBlockAsm12B: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBlockAsm12B - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeSnappyBlockAsm12B - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeSnappyBlockAsm12B: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBlockAsm12B JB match_nolit_end_encodeSnappyBlockAsm12B - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeSnappyBlockAsm12B - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeSnappyBlockAsm12B matchlen_match1_match_nolit_encodeSnappyBlockAsm12B: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeSnappyBlockAsm12B - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeSnappyBlockAsm12B: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBlockAsm12B: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBlockAsm12B - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBlockAsm12B two_byte_offset_short_match_nolit_encodeSnappyBlockAsm12B: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm12B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm12B - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBlockAsm12B emit_copy_three_match_nolit_encodeSnappyBlockAsm12B: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBlockAsm12B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm12B - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBlockAsm12B: - MOVQ $0x000000cf1bbcdcbb, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x18, DI - IMULQ R8, DI - SHRQ $0x34, DI - SHLQ $0x18, BX - IMULQ R8, BX - SHRQ $0x34, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x000000cf1bbcdcbb, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x18, R8 + IMULQ R9, R8 + SHRQ $0x34, R8 + SHLQ $0x18, SI + IMULQ R9, SI + SHRQ $0x34, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeSnappyBlockAsm12B - INCL CX + INCL DX JMP search_loop_encodeSnappyBlockAsm12B emit_remainder_encodeSnappyBlockAsm12B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBlockAsm12B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBlockAsm12B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBlockAsm12B @@ -13181,26 +13194,26 @@ emit_remainder_ok_encodeSnappyBlockAsm12B: JB three_bytes_emit_remainder_encodeSnappyBlockAsm12B three_bytes_emit_remainder_encodeSnappyBlockAsm12B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm12B two_bytes_emit_remainder_encodeSnappyBlockAsm12B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBlockAsm12B JMP memmove_long_emit_remainder_encodeSnappyBlockAsm12B one_byte_emit_remainder_encodeSnappyBlockAsm12B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -13216,73 +13229,73 @@ memmove_emit_remainder_encodeSnappyBlockAsm12B: JMP emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm12B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBlockAsm12B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBlockAsm12B memmove_long_emit_remainder_encodeSnappyBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm12Blarge_big_loop_back: MOVOU (SI), X4 @@ -13296,718 +13309,719 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm12Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBlockAsm12B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBlockAsm10B(dst []byte, src []byte) int +// func encodeSnappyBlockAsm10B(dst []byte, src []byte, tmp *[4096]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBlockAsm10B(SB), $4120-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000020, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBlockAsm10B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000020, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBlockAsm10B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBlockAsm10B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBlockAsm10B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm10B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x9e3779b1, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x9e3779b1, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x20, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x36, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x20, R11 + IMULQ R9, R11 + SHRQ $0x36, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x20, R10 + IMULQ R9, R10 + SHRQ $0x36, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeSnappyBlockAsm10B - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_encodeSnappyBlockAsm10B repeat_extend_back_loop_encodeSnappyBlockAsm10B: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_encodeSnappyBlockAsm10B - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeSnappyBlockAsm10B - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_encodeSnappyBlockAsm10B repeat_extend_back_end_encodeSnappyBlockAsm10B: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeSnappyBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeSnappyBlockAsm10B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeSnappyBlockAsm10B - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeSnappyBlockAsm10B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeSnappyBlockAsm10B JB three_bytes_repeat_emit_encodeSnappyBlockAsm10B three_bytes_repeat_emit_encodeSnappyBlockAsm10B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm10B two_bytes_repeat_emit_encodeSnappyBlockAsm10B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeSnappyBlockAsm10B JMP memmove_long_repeat_emit_encodeSnappyBlockAsm10B one_byte_repeat_emit_encodeSnappyBlockAsm10B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeSnappyBlockAsm10B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveShort - CMPQ DI, $0x08 + CMPQ R8, $0x08 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_8 - CMPQ DI, $0x10 + CMPQ R8, $0x10 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_8through16 - CMPQ DI, $0x20 + CMPQ R8, $0x20 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_8: - MOVQ (R8), R9 - MOVQ R9, (AX) + MOVQ (R9), R10 + MOVQ R10, (CX) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm10B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_8through16: - MOVQ (R8), R9 - MOVQ -8(R8)(DI*1), R8 - MOVQ R9, (AX) - MOVQ R8, -8(AX)(DI*1) + MOVQ (R9), R10 + MOVQ -8(R9)(R8*1), R9 + MOVQ R10, (CX) + MOVQ R9, -8(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm10B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_17through32: - MOVOU (R8), X0 - MOVOU -16(R8)(DI*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU -16(R9)(R8*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm10B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm10B_memmove_move_33through64: - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) memmove_end_copy_repeat_emit_encodeSnappyBlockAsm10B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeSnappyBlockAsm10B memmove_long_repeat_emit_encodeSnappyBlockAsm10B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveLong - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVQ DI, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVQ R8, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(R8)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(R9)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm10Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(R8)(R11*1), X4 - MOVOU -16(R8)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ DI, R11 + MOVOU -32(R9)(R12*1), X4 + MOVOU -16(R9)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R8, R12 JAE emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeSnappyBlockAsm10B: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm10B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_encodeSnappyBlockAsm10B - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm10B - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm10B - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm10B matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm10B matchlen_match8_repeat_extend_encodeSnappyBlockAsm10B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_encodeSnappyBlockAsm10B - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm10B - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_encodeSnappyBlockAsm10B matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm10B matchlen_match4_repeat_extend_encodeSnappyBlockAsm10B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_encodeSnappyBlockAsm10B - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_encodeSnappyBlockAsm10B - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_encodeSnappyBlockAsm10B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_encodeSnappyBlockAsm10B JB repeat_extend_forward_end_encodeSnappyBlockAsm10B - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_encodeSnappyBlockAsm10B - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_encodeSnappyBlockAsm10B matchlen_match1_repeat_extend_encodeSnappyBlockAsm10B: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_encodeSnappyBlockAsm10B - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_encodeSnappyBlockAsm10B: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm10B: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm10B - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm10B two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm10B: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm10B - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm10B - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeSnappyBlockAsm10B emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm10B: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeSnappyBlockAsm10B: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeSnappyBlockAsm10B no_repeat_found_encodeSnappyBlockAsm10B: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBlockAsm10B - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeSnappyBlockAsm10B - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeSnappyBlockAsm10B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBlockAsm10B candidate3_match_encodeSnappyBlockAsm10B: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeSnappyBlockAsm10B candidate2_match_encodeSnappyBlockAsm10B: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeSnappyBlockAsm10B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBlockAsm10B match_extend_back_loop_encodeSnappyBlockAsm10B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBlockAsm10B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBlockAsm10B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBlockAsm10B JMP match_extend_back_loop_encodeSnappyBlockAsm10B match_extend_back_end_encodeSnappyBlockAsm10B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBlockAsm10B: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeSnappyBlockAsm10B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeSnappyBlockAsm10B - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeSnappyBlockAsm10B JB three_bytes_match_emit_encodeSnappyBlockAsm10B three_bytes_match_emit_encodeSnappyBlockAsm10B: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm10B two_bytes_match_emit_encodeSnappyBlockAsm10B: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeSnappyBlockAsm10B JMP memmove_long_match_emit_encodeSnappyBlockAsm10B one_byte_match_emit_encodeSnappyBlockAsm10B: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBlockAsm10B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm10B emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm10B emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm10B emit_lit_memmove_match_emit_encodeSnappyBlockAsm10B_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBlockAsm10B: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeSnappyBlockAsm10B memmove_long_match_emit_encodeSnappyBlockAsm10B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm10Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeSnappyBlockAsm10B: match_nolit_loop_encodeSnappyBlockAsm10B: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm10B: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBlockAsm10B - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm10B - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeSnappyBlockAsm10B - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm10B matchlen_bsf_16match_nolit_encodeSnappyBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm10B matchlen_match8_match_nolit_encodeSnappyBlockAsm10B: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBlockAsm10B - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm10B - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeSnappyBlockAsm10B matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm10B matchlen_match4_match_nolit_encodeSnappyBlockAsm10B: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBlockAsm10B - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeSnappyBlockAsm10B - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeSnappyBlockAsm10B: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBlockAsm10B JB match_nolit_end_encodeSnappyBlockAsm10B - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeSnappyBlockAsm10B - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeSnappyBlockAsm10B matchlen_match1_match_nolit_encodeSnappyBlockAsm10B: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeSnappyBlockAsm10B - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeSnappyBlockAsm10B: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBlockAsm10B: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBlockAsm10B - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBlockAsm10B two_byte_offset_short_match_nolit_encodeSnappyBlockAsm10B: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm10B - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm10B - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBlockAsm10B emit_copy_three_match_nolit_encodeSnappyBlockAsm10B: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBlockAsm10B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm10B - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBlockAsm10B: - MOVQ $0x9e3779b1, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x20, DI - IMULQ R8, DI - SHRQ $0x36, DI - SHLQ $0x20, BX - IMULQ R8, BX - SHRQ $0x36, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x9e3779b1, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x20, R8 + IMULQ R9, R8 + SHRQ $0x36, R8 + SHLQ $0x20, SI + IMULQ R9, SI + SHRQ $0x36, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeSnappyBlockAsm10B - INCL CX + INCL DX JMP search_loop_encodeSnappyBlockAsm10B emit_remainder_encodeSnappyBlockAsm10B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBlockAsm10B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBlockAsm10B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBlockAsm10B @@ -14016,26 +14030,26 @@ emit_remainder_ok_encodeSnappyBlockAsm10B: JB three_bytes_emit_remainder_encodeSnappyBlockAsm10B three_bytes_emit_remainder_encodeSnappyBlockAsm10B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm10B two_bytes_emit_remainder_encodeSnappyBlockAsm10B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBlockAsm10B JMP memmove_long_emit_remainder_encodeSnappyBlockAsm10B one_byte_emit_remainder_encodeSnappyBlockAsm10B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -14051,73 +14065,73 @@ memmove_emit_remainder_encodeSnappyBlockAsm10B: JMP emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm10B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBlockAsm10B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBlockAsm10B memmove_long_emit_remainder_encodeSnappyBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm10Blarge_big_loop_back: MOVOU (SI), X4 @@ -14131,714 +14145,715 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm10Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBlockAsm10B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBlockAsm8B(dst []byte, src []byte) int +// func encodeSnappyBlockAsm8B(dst []byte, src []byte, tmp *[1024]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBlockAsm8B(SB), $1048-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000008, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBlockAsm8B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000008, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBlockAsm8B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBlockAsm8B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBlockAsm8B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x04, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x04, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm8B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x9e3779b1, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x38, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x9e3779b1, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x20, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x38, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x38, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x20, R11 + IMULQ R9, R11 + SHRQ $0x38, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x20, R10 + IMULQ R9, R10 + SHRQ $0x38, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_encodeSnappyBlockAsm8B - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_encodeSnappyBlockAsm8B repeat_extend_back_loop_encodeSnappyBlockAsm8B: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_encodeSnappyBlockAsm8B - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_encodeSnappyBlockAsm8B - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_encodeSnappyBlockAsm8B repeat_extend_back_end_encodeSnappyBlockAsm8B: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_encodeSnappyBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET repeat_dst_size_check_encodeSnappyBlockAsm8B: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_encodeSnappyBlockAsm8B - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_encodeSnappyBlockAsm8B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_encodeSnappyBlockAsm8B JB three_bytes_repeat_emit_encodeSnappyBlockAsm8B three_bytes_repeat_emit_encodeSnappyBlockAsm8B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_repeat_emit_encodeSnappyBlockAsm8B two_bytes_repeat_emit_encodeSnappyBlockAsm8B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_encodeSnappyBlockAsm8B JMP memmove_long_repeat_emit_encodeSnappyBlockAsm8B one_byte_repeat_emit_encodeSnappyBlockAsm8B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_repeat_emit_encodeSnappyBlockAsm8B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveShort - CMPQ DI, $0x08 + CMPQ R8, $0x08 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_8 - CMPQ DI, $0x10 + CMPQ R8, $0x10 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_8through16 - CMPQ DI, $0x20 + CMPQ R8, $0x20 JBE emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_33through64 emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_8: - MOVQ (R8), R9 - MOVQ R9, (AX) + MOVQ (R9), R10 + MOVQ R10, (CX) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm8B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_8through16: - MOVQ (R8), R9 - MOVQ -8(R8)(DI*1), R8 - MOVQ R9, (AX) - MOVQ R8, -8(AX)(DI*1) + MOVQ (R9), R10 + MOVQ -8(R9)(R8*1), R9 + MOVQ R10, (CX) + MOVQ R9, -8(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm8B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_17through32: - MOVOU (R8), X0 - MOVOU -16(R8)(DI*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU -16(R9)(R8*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R8*1) JMP memmove_end_copy_repeat_emit_encodeSnappyBlockAsm8B emit_lit_memmove_repeat_emit_encodeSnappyBlockAsm8B_memmove_move_33through64: - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) memmove_end_copy_repeat_emit_encodeSnappyBlockAsm8B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_repeat_emit_encodeSnappyBlockAsm8B memmove_long_repeat_emit_encodeSnappyBlockAsm8B: - LEAQ (AX)(DI*1), BX + LEAQ (CX)(R8*1), SI // genMemMoveLong - MOVOU (R8), X0 - MOVOU 16(R8), X1 - MOVOU -32(R8)(DI*1), X2 - MOVOU -16(R8)(DI*1), X3 - MOVQ DI, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (R9), X0 + MOVOU 16(R9), X1 + MOVOU -32(R9)(R8*1), X2 + MOVOU -16(R9)(R8*1), X3 + MOVQ R8, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(R8)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(R9)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm8Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(R8)(R11*1), X4 - MOVOU -16(R8)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ DI, R11 + MOVOU -32(R9)(R12*1), X4 + MOVOU -16(R9)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R8, R12 JAE emit_lit_memmove_long_repeat_emit_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(DI*1) - MOVOU X3, -16(AX)(DI*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R8*1) + MOVOU X3, -16(CX)(R8*1) + MOVQ SI, CX emit_literal_done_repeat_emit_encodeSnappyBlockAsm8B: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm8B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_encodeSnappyBlockAsm8B - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm8B - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm8B - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_encodeSnappyBlockAsm8B matchlen_bsf_16repeat_extend_encodeSnappyBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm8B matchlen_match8_repeat_extend_encodeSnappyBlockAsm8B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_encodeSnappyBlockAsm8B - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm8B - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_encodeSnappyBlockAsm8B matchlen_bsf_8_repeat_extend_encodeSnappyBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_encodeSnappyBlockAsm8B matchlen_match4_repeat_extend_encodeSnappyBlockAsm8B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_encodeSnappyBlockAsm8B - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_encodeSnappyBlockAsm8B - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_encodeSnappyBlockAsm8B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_encodeSnappyBlockAsm8B JB repeat_extend_forward_end_encodeSnappyBlockAsm8B - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_encodeSnappyBlockAsm8B - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_encodeSnappyBlockAsm8B matchlen_match1_repeat_extend_encodeSnappyBlockAsm8B: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_encodeSnappyBlockAsm8B - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_encodeSnappyBlockAsm8B: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm8B: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm8B - MOVB $0xee, (AX) - MOVW SI, 1(AX) - LEAL -60(BX), BX - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW DI, 1(CX) + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_encodeSnappyBlockAsm8B two_byte_offset_short_repeat_as_copy_encodeSnappyBlockAsm8B: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm8B - LEAL -15(DI), DI - MOVB SI, 1(AX) - SHRL $0x08, SI - SHLL $0x05, SI - ORL SI, DI - MOVB DI, (AX) - ADDQ $0x02, AX + LEAL -15(R8), R8 + MOVB DI, 1(CX) + SHRL $0x08, DI + SHLL $0x05, DI + ORL DI, R8 + MOVB R8, (CX) + ADDQ $0x02, CX JMP repeat_end_emit_encodeSnappyBlockAsm8B emit_copy_three_repeat_as_copy_encodeSnappyBlockAsm8B: - LEAL -2(DI), DI - MOVB DI, (AX) - MOVW SI, 1(AX) - ADDQ $0x03, AX + LEAL -2(R8), R8 + MOVB R8, (CX) + MOVW DI, 1(CX) + ADDQ $0x03, CX repeat_end_emit_encodeSnappyBlockAsm8B: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_encodeSnappyBlockAsm8B no_repeat_found_encodeSnappyBlockAsm8B: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBlockAsm8B - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_encodeSnappyBlockAsm8B - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_encodeSnappyBlockAsm8B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBlockAsm8B candidate3_match_encodeSnappyBlockAsm8B: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_encodeSnappyBlockAsm8B candidate2_match_encodeSnappyBlockAsm8B: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_encodeSnappyBlockAsm8B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBlockAsm8B match_extend_back_loop_encodeSnappyBlockAsm8B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBlockAsm8B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBlockAsm8B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBlockAsm8B JMP match_extend_back_loop_encodeSnappyBlockAsm8B match_extend_back_end_encodeSnappyBlockAsm8B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBlockAsm8B: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_encodeSnappyBlockAsm8B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), DI - CMPL DI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), R8 + CMPL R8, $0x3c JB one_byte_match_emit_encodeSnappyBlockAsm8B - CMPL DI, $0x00000100 + CMPL R8, $0x00000100 JB two_bytes_match_emit_encodeSnappyBlockAsm8B JB three_bytes_match_emit_encodeSnappyBlockAsm8B three_bytes_match_emit_encodeSnappyBlockAsm8B: - MOVB $0xf4, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBlockAsm8B two_bytes_match_emit_encodeSnappyBlockAsm8B: - MOVB $0xf0, (AX) - MOVB DI, 1(AX) - ADDQ $0x02, AX - CMPL DI, $0x40 + MOVB $0xf0, (CX) + MOVB R8, 1(CX) + ADDQ $0x02, CX + CMPL R8, $0x40 JB memmove_match_emit_encodeSnappyBlockAsm8B JMP memmove_long_match_emit_encodeSnappyBlockAsm8B one_byte_match_emit_encodeSnappyBlockAsm8B: - SHLB $0x02, DI - MOVB DI, (AX) - ADDQ $0x01, AX + SHLB $0x02, R8 + MOVB R8, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBlockAsm8B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_8: - MOVQ (SI), R9 - MOVQ R9, (AX) + MOVQ (DI), R10 + MOVQ R10, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm8B emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_8through16: - MOVQ (SI), R9 - MOVQ -8(SI)(R8*1), SI - MOVQ R9, (AX) - MOVQ SI, -8(AX)(R8*1) + MOVQ (DI), R10 + MOVQ -8(DI)(R9*1), DI + MOVQ R10, (CX) + MOVQ DI, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm8B emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_17through32: - MOVOU (SI), X0 - MOVOU -16(SI)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU -16(DI)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBlockAsm8B emit_lit_memmove_match_emit_encodeSnappyBlockAsm8B_memmove_move_33through64: - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBlockAsm8B: - MOVQ DI, AX + MOVQ R8, CX JMP emit_literal_done_match_emit_encodeSnappyBlockAsm8B memmove_long_match_emit_encodeSnappyBlockAsm8B: - LEAQ (AX)(R8*1), DI + LEAQ (CX)(R9*1), R8 // genMemMoveLong - MOVOU (SI), X0 - MOVOU 16(SI), X1 - MOVOU -32(SI)(R8*1), X2 - MOVOU -16(SI)(R8*1), X3 - MOVQ R8, R10 - SHRQ $0x05, R10 - MOVQ AX, R9 - ANDL $0x0000001f, R9 - MOVQ $0x00000040, R11 - SUBQ R9, R11 - DECQ R10 + MOVOU (DI), X0 + MOVOU 16(DI), X1 + MOVOU -32(DI)(R9*1), X2 + MOVOU -16(DI)(R9*1), X3 + MOVQ R9, R11 + SHRQ $0x05, R11 + MOVQ CX, R10 + ANDL $0x0000001f, R10 + MOVQ $0x00000040, R12 + SUBQ R10, R12 + DECQ R11 JA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(SI)(R11*1), R9 - LEAQ -32(AX)(R11*1), R12 + LEAQ -32(DI)(R12*1), R10 + LEAQ -32(CX)(R12*1), R13 emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm8Blarge_big_loop_back: - MOVOU (R9), X4 - MOVOU 16(R9), X5 - MOVOA X4, (R12) - MOVOA X5, 16(R12) + MOVOU (R10), X4 + MOVOU 16(R10), X5 + MOVOA X4, (R13) + MOVOA X5, 16(R13) + ADDQ $0x20, R13 + ADDQ $0x20, R10 ADDQ $0x20, R12 - ADDQ $0x20, R9 - ADDQ $0x20, R11 - DECQ R10 + DECQ R11 JNA emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(SI)(R11*1), X4 - MOVOU -16(SI)(R11*1), X5 - MOVOA X4, -32(AX)(R11*1) - MOVOA X5, -16(AX)(R11*1) - ADDQ $0x20, R11 - CMPQ R8, R11 + MOVOU -32(DI)(R12*1), X4 + MOVOU -16(DI)(R12*1), X5 + MOVOA X4, -32(CX)(R12*1) + MOVOA X5, -16(CX)(R12*1) + ADDQ $0x20, R12 + CMPQ R9, R12 JAE emit_lit_memmove_long_match_emit_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ DI, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ R8, CX emit_literal_done_match_emit_encodeSnappyBlockAsm8B: match_nolit_loop_encodeSnappyBlockAsm8B: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm8B: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBlockAsm8B - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm8B - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_encodeSnappyBlockAsm8B - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_encodeSnappyBlockAsm8B matchlen_bsf_16match_nolit_encodeSnappyBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm8B matchlen_match8_match_nolit_encodeSnappyBlockAsm8B: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBlockAsm8B - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm8B - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_encodeSnappyBlockAsm8B matchlen_bsf_8_match_nolit_encodeSnappyBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_encodeSnappyBlockAsm8B matchlen_match4_match_nolit_encodeSnappyBlockAsm8B: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBlockAsm8B - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_encodeSnappyBlockAsm8B - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_encodeSnappyBlockAsm8B: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBlockAsm8B JB match_nolit_end_encodeSnappyBlockAsm8B - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_encodeSnappyBlockAsm8B - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_encodeSnappyBlockAsm8B matchlen_match1_match_nolit_encodeSnappyBlockAsm8B: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_encodeSnappyBlockAsm8B - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_encodeSnappyBlockAsm8B: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBlockAsm8B: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBlockAsm8B - MOVB $0xee, (AX) - MOVW BX, 1(AX) - LEAL -60(R9), R9 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW SI, 1(CX) + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBlockAsm8B two_byte_offset_short_match_nolit_encodeSnappyBlockAsm8B: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBlockAsm8B - LEAL -15(SI), SI - MOVB BL, 1(AX) - SHRL $0x08, BX - SHLL $0x05, BX - ORL BX, SI - MOVB SI, (AX) - ADDQ $0x02, AX + LEAL -15(DI), DI + MOVB SI, 1(CX) + SHRL $0x08, SI + SHLL $0x05, SI + ORL SI, DI + MOVB DI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBlockAsm8B emit_copy_three_match_nolit_encodeSnappyBlockAsm8B: - LEAL -2(SI), SI - MOVB SI, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + LEAL -2(DI), DI + MOVB DI, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBlockAsm8B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBlockAsm8B - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBlockAsm8B: - MOVQ $0x9e3779b1, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x20, DI - IMULQ R8, DI - SHRQ $0x38, DI - SHLQ $0x20, BX - IMULQ R8, BX - SHRQ $0x38, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x9e3779b1, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x20, R8 + IMULQ R9, R8 + SHRQ $0x38, R8 + SHLQ $0x20, SI + IMULQ R9, SI + SHRQ $0x38, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_encodeSnappyBlockAsm8B - INCL CX + INCL DX JMP search_loop_encodeSnappyBlockAsm8B emit_remainder_encodeSnappyBlockAsm8B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBlockAsm8B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBlockAsm8B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBlockAsm8B @@ -14847,26 +14862,26 @@ emit_remainder_ok_encodeSnappyBlockAsm8B: JB three_bytes_emit_remainder_encodeSnappyBlockAsm8B three_bytes_emit_remainder_encodeSnappyBlockAsm8B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBlockAsm8B two_bytes_emit_remainder_encodeSnappyBlockAsm8B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBlockAsm8B JMP memmove_long_emit_remainder_encodeSnappyBlockAsm8B one_byte_emit_remainder_encodeSnappyBlockAsm8B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -14882,73 +14897,73 @@ memmove_emit_remainder_encodeSnappyBlockAsm8B: JMP emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBlockAsm8B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBlockAsm8B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBlockAsm8B memmove_long_emit_remainder_encodeSnappyBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm8Blarge_big_loop_back: MOVOU (SI), X4 @@ -14962,520 +14977,521 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm8Blarge_big_loop_back: JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBlockAsm8B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBetterBlockAsm(dst []byte, src []byte) int +// func encodeSnappyBetterBlockAsm(dst []byte, src []byte, tmp *[589824]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBetterBlockAsm(SB), $589848-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00001200, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBetterBlockAsm(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00001200, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBetterBlockAsm: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBetterBlockAsm MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBetterBlockAsm: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x07, BX - CMPL BX, $0x63 + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x07, SI + CMPL SI, $0x63 JBE check_maxskip_ok_encodeSnappyBetterBlockAsm - LEAL 100(CX), BX + LEAL 100(DX), SI JMP check_maxskip_cont_encodeSnappyBetterBlockAsm check_maxskip_ok_encodeSnappyBetterBlockAsm: - LEAL 1(CX)(BX*1), BX + LEAL 1(DX)(SI*1), SI check_maxskip_cont_encodeSnappyBetterBlockAsm: - CMPL BX, 8(SP) + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x00cf1bbcdcbfa563, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x2f, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 524312(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 524312(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x00cf1bbcdcbfa563, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x2f, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x32, R11 + MOVL (AX)(R10*4), SI + MOVL 524288(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 524288(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeSnappyBetterBlockAsm - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeSnappyBetterBlockAsm no_short_found_encodeSnappyBetterBlockAsm: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeSnappyBetterBlockAsm - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBetterBlockAsm candidateS_match_encodeSnappyBetterBlockAsm: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x2f, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x2f, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBetterBlockAsm - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeSnappyBetterBlockAsm: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm match_extend_back_loop_encodeSnappyBetterBlockAsm: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBetterBlockAsm - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBetterBlockAsm - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm JMP match_extend_back_loop_encodeSnappyBetterBlockAsm match_extend_back_end_encodeSnappyBetterBlockAsm: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 5(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 5(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBetterBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBetterBlockAsm: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm JB match_nolit_end_encodeSnappyBetterBlockAsm - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeSnappyBetterBlockAsm matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeSnappyBetterBlockAsm - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeSnappyBetterBlockAsm: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - CMPL R11, $0x01 + CMPL R12, $0x01 JA match_length_ok_encodeSnappyBetterBlockAsm - CMPL DI, $0x0000ffff + CMPL R8, $0x0000ffff JBE match_length_ok_encodeSnappyBetterBlockAsm - MOVL 20(SP), CX - INCL CX + MOVL 20(SP), DX + INCL DX JMP search_loop_encodeSnappyBetterBlockAsm match_length_ok_encodeSnappyBetterBlockAsm: - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeSnappyBetterBlockAsm - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeSnappyBetterBlockAsm - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeSnappyBetterBlockAsm - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_match_emit_encodeSnappyBetterBlockAsm - CMPL BX, $0x01000000 + CMPL SI, $0x01000000 JB four_bytes_match_emit_encodeSnappyBetterBlockAsm - MOVB $0xfc, (AX) - MOVL BX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL SI, 1(CX) + ADDQ $0x05, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm four_bytes_match_emit_encodeSnappyBetterBlockAsm: - MOVL BX, R10 - SHRL $0x10, R10 - MOVB $0xf8, (AX) - MOVW BX, 1(AX) - MOVB R10, 3(AX) - ADDQ $0x04, AX + MOVL SI, R11 + SHRL $0x10, R11 + MOVB $0xf8, (CX) + MOVW SI, 1(CX) + MOVB R11, 3(CX) + ADDQ $0x04, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm three_bytes_match_emit_encodeSnappyBetterBlockAsm: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm two_bytes_match_emit_encodeSnappyBetterBlockAsm: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeSnappyBetterBlockAsm JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm one_byte_match_emit_encodeSnappyBetterBlockAsm: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBetterBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeSnappyBetterBlockAsm memmove_long_match_emit_encodeSnappyBetterBlockAsm: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsmlarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsmlarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeSnappyBetterBlockAsm: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy - CMPL DI, $0x00010000 + CMPL R8, $0x00010000 JB two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm four_bytes_loop_back_match_nolit_encodeSnappyBetterBlockAsm: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE four_bytes_remain_match_nolit_encodeSnappyBetterBlockAsm - MOVB $0xff, (AX) - MOVL DI, 1(AX) - LEAL -64(R11), R11 - ADDQ $0x05, AX - CMPL R11, $0x04 + MOVB $0xff, (CX) + MOVL R8, 1(CX) + LEAL -64(R12), R12 + ADDQ $0x05, CX + CMPL R12, $0x04 JB four_bytes_remain_match_nolit_encodeSnappyBetterBlockAsm JMP four_bytes_loop_back_match_nolit_encodeSnappyBetterBlockAsm four_bytes_remain_match_nolit_encodeSnappyBetterBlockAsm: - TESTL R11, R11 + TESTL R12, R12 JZ match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm - XORL BX, BX - LEAL -1(BX)(R11*4), R11 - MOVB R11, (AX) - MOVL DI, 1(AX) - ADDQ $0x05, AX + XORL SI, SI + LEAL -1(SI)(R12*4), R12 + MOVB R12, (CX) + MOVL R8, 1(CX) + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBetterBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBetterBlockAsm: - MOVQ $0x00cf1bbcdcbfa563, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x2f, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x32, R10 - SHLQ $0x08, R11 - IMULQ BX, R11 - SHRQ $0x2f, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x32, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 524312(SP)(R10*4) - MOVL R13, 524312(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x00cf1bbcdcbfa563, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x08, R10 + IMULQ SI, R10 + SHRQ $0x2f, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x32, R11 + SHLQ $0x08, R12 + IMULQ SI, R12 + SHRQ $0x2f, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x32, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 524288(AX)(R11*4) + MOVL R14, 524288(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeSnappyBetterBlockAsm: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeSnappyBetterBlockAsm - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x2f, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x08, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x2f, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x08, R11 + IMULQ SI, R11 + SHRQ $0x2f, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeSnappyBetterBlockAsm emit_remainder_encodeSnappyBetterBlockAsm: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 5(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 5(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBetterBlockAsm - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBetterBlockAsm: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBetterBlockAsm @@ -15485,41 +15501,41 @@ emit_remainder_ok_encodeSnappyBetterBlockAsm: JB three_bytes_emit_remainder_encodeSnappyBetterBlockAsm CMPL DX, $0x01000000 JB four_bytes_emit_remainder_encodeSnappyBetterBlockAsm - MOVB $0xfc, (AX) - MOVL DX, 1(AX) - ADDQ $0x05, AX + MOVB $0xfc, (CX) + MOVL DX, 1(CX) + ADDQ $0x05, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm four_bytes_emit_remainder_encodeSnappyBetterBlockAsm: MOVL DX, BX SHRL $0x10, BX - MOVB $0xf8, (AX) - MOVW DX, 1(AX) - MOVB BL, 3(AX) - ADDQ $0x04, AX + MOVB $0xf8, (CX) + MOVW DX, 1(CX) + MOVB BL, 3(CX) + ADDQ $0x04, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm three_bytes_emit_remainder_encodeSnappyBetterBlockAsm: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm two_bytes_emit_remainder_encodeSnappyBetterBlockAsm: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBetterBlockAsm JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm one_byte_emit_remainder_encodeSnappyBetterBlockAsm: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBetterBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -15535,73 +15551,73 @@ memmove_emit_remainder_encodeSnappyBetterBlockAsm: JMP emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm memmove_long_emit_remainder_encodeSnappyBetterBlockAsm: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsmlarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsmlarge_big_loop_back: MOVOU (SI), X4 @@ -15615,463 +15631,464 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsmlarge_big_loop_ba JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsmlarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsmlarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsmlarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBetterBlockAsm64K(dst []byte, src []byte) int +// func encodeSnappyBetterBlockAsm64K(dst []byte, src []byte, tmp *[294912]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBetterBlockAsm64K(SB), $327704-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000a00, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBetterBlockAsm64K(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000900, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBetterBlockAsm64K: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBetterBlockAsm64K MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBetterBlockAsm64K: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x07, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x07, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm64K - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x00cf1bbcdcbfa563, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x30, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x32, R10 - MOVL 24(SP)(R9*4), BX - MOVL 262168(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 262168(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x00cf1bbcdcbfa563, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x30, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x33, R11 + MOVL (AX)(R10*4), SI + MOVL 262144(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 262144(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm64K - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeSnappyBetterBlockAsm64K - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeSnappyBetterBlockAsm64K no_short_found_encodeSnappyBetterBlockAsm64K: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm64K - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeSnappyBetterBlockAsm64K - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBetterBlockAsm64K candidateS_match_encodeSnappyBetterBlockAsm64K: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x08, R9 - IMULQ R8, R9 - SHRQ $0x30, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x08, R10 + IMULQ R9, R10 + SHRQ $0x30, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBetterBlockAsm64K - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeSnappyBetterBlockAsm64K: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm64K match_extend_back_loop_encodeSnappyBetterBlockAsm64K: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBetterBlockAsm64K - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBetterBlockAsm64K - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm64K JMP match_extend_back_loop_encodeSnappyBetterBlockAsm64K match_extend_back_end_encodeSnappyBetterBlockAsm64K: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBetterBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBetterBlockAsm64K: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm64K: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm64K - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm64K - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm64K - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm64K matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm64K: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm64K matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm64K: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm64K - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm64K - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm64K matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm64K: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm64K matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm64K: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm64K - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm64K - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm64K: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm64K JB match_nolit_end_encodeSnappyBetterBlockAsm64K - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm64K - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeSnappyBetterBlockAsm64K matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm64K: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeSnappyBetterBlockAsm64K - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeSnappyBetterBlockAsm64K: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeSnappyBetterBlockAsm64K - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeSnappyBetterBlockAsm64K - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeSnappyBetterBlockAsm64K JB three_bytes_match_emit_encodeSnappyBetterBlockAsm64K three_bytes_match_emit_encodeSnappyBetterBlockAsm64K: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm64K two_bytes_match_emit_encodeSnappyBetterBlockAsm64K: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeSnappyBetterBlockAsm64K JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm64K one_byte_match_emit_encodeSnappyBetterBlockAsm64K: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBetterBlockAsm64K: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm64K emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm64K emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm64K emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm64K_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm64K: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeSnappyBetterBlockAsm64K memmove_long_match_emit_encodeSnappyBetterBlockAsm64K: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm64Klarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm64Klarge_big_loop_back - -emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + +emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32: + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeSnappyBetterBlockAsm64K: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm64K: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm64K - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm64K two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm64K: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm64K - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm64K - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm64K emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm64K: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm64K: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm64K - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBetterBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBetterBlockAsm64K: - MOVQ $0x00cf1bbcdcbfa563, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x30, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x32, R10 - SHLQ $0x08, R11 - IMULQ BX, R11 - SHRQ $0x30, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x32, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 262168(SP)(R10*4) - MOVL R13, 262168(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x00cf1bbcdcbfa563, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x08, R10 + IMULQ SI, R10 + SHRQ $0x30, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x33, R11 + SHLQ $0x08, R12 + IMULQ SI, R12 + SHRQ $0x30, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x33, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 262144(AX)(R11*4) + MOVL R14, 262144(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeSnappyBetterBlockAsm64K: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeSnappyBetterBlockAsm64K - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x08, R9 - IMULQ BX, R9 - SHRQ $0x30, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x08, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x30, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x08, R11 + IMULQ SI, R11 + SHRQ $0x30, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeSnappyBetterBlockAsm64K emit_remainder_encodeSnappyBetterBlockAsm64K: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBetterBlockAsm64K - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBetterBlockAsm64K: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm64K - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBetterBlockAsm64K @@ -16080,26 +16097,26 @@ emit_remainder_ok_encodeSnappyBetterBlockAsm64K: JB three_bytes_emit_remainder_encodeSnappyBetterBlockAsm64K three_bytes_emit_remainder_encodeSnappyBetterBlockAsm64K: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64K two_bytes_emit_remainder_encodeSnappyBetterBlockAsm64K: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBetterBlockAsm64K JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64K one_byte_emit_remainder_encodeSnappyBetterBlockAsm64K: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBetterBlockAsm64K: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -16115,73 +16132,73 @@ memmove_emit_remainder_encodeSnappyBetterBlockAsm64K: JMP emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm64K emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm64K_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm64K: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm64K memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64K: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64Klarge_big_loop_back: MOVOU (SI), X4 @@ -16195,463 +16212,464 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64Klarge_big_loop JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64Klarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm64Klarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm64K: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBetterBlockAsm12B(dst []byte, src []byte) int +// func encodeSnappyBetterBlockAsm12B(dst []byte, src []byte, tmp *[81920]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBetterBlockAsm12B(SB), $81944-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000280, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBetterBlockAsm12B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000280, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBetterBlockAsm12B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBetterBlockAsm12B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBetterBlockAsm12B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x06, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x06, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm12B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x34, R10 - MOVL 24(SP)(R9*4), BX - MOVL 65560(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 65560(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x34, R11 + MOVL (AX)(R10*4), SI + MOVL 65536(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 65536(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm12B - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeSnappyBetterBlockAsm12B - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeSnappyBetterBlockAsm12B no_short_found_encodeSnappyBetterBlockAsm12B: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm12B - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeSnappyBetterBlockAsm12B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBetterBlockAsm12B candidateS_match_encodeSnappyBetterBlockAsm12B: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x32, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x32, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBetterBlockAsm12B - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeSnappyBetterBlockAsm12B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm12B match_extend_back_loop_encodeSnappyBetterBlockAsm12B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBetterBlockAsm12B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBetterBlockAsm12B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm12B JMP match_extend_back_loop_encodeSnappyBetterBlockAsm12B match_extend_back_end_encodeSnappyBetterBlockAsm12B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBetterBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBetterBlockAsm12B: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm12B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm12B - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm12B - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm12B - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm12B matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm12B matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm12B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm12B - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm12B - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm12B matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm12B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm12B matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm12B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm12B - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm12B - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm12B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm12B JB match_nolit_end_encodeSnappyBetterBlockAsm12B - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm12B - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeSnappyBetterBlockAsm12B matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm12B: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeSnappyBetterBlockAsm12B - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeSnappyBetterBlockAsm12B: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeSnappyBetterBlockAsm12B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeSnappyBetterBlockAsm12B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeSnappyBetterBlockAsm12B JB three_bytes_match_emit_encodeSnappyBetterBlockAsm12B three_bytes_match_emit_encodeSnappyBetterBlockAsm12B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm12B two_bytes_match_emit_encodeSnappyBetterBlockAsm12B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeSnappyBetterBlockAsm12B JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm12B one_byte_match_emit_encodeSnappyBetterBlockAsm12B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBetterBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm12B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm12B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm12B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm12B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm12B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeSnappyBetterBlockAsm12B memmove_long_match_emit_encodeSnappyBetterBlockAsm12B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm12Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeSnappyBetterBlockAsm12B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm12B: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm12B - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm12B two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm12B: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm12B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm12B - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm12B emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm12B: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm12B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm12B - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBetterBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBetterBlockAsm12B: - MOVQ $0x0000cf1bbcdcbf9b, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x32, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x34, R10 - SHLQ $0x10, R11 - IMULQ BX, R11 - SHRQ $0x32, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x34, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 65560(SP)(R10*4) - MOVL R13, 65560(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x0000cf1bbcdcbf9b, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x10, R10 + IMULQ SI, R10 + SHRQ $0x32, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x34, R11 + SHLQ $0x10, R12 + IMULQ SI, R12 + SHRQ $0x32, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x34, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 65536(AX)(R11*4) + MOVL R14, 65536(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeSnappyBetterBlockAsm12B: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeSnappyBetterBlockAsm12B - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x32, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x10, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x32, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x10, R11 + IMULQ SI, R11 + SHRQ $0x32, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeSnappyBetterBlockAsm12B emit_remainder_encodeSnappyBetterBlockAsm12B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBetterBlockAsm12B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBetterBlockAsm12B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm12B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBetterBlockAsm12B @@ -16660,26 +16678,26 @@ emit_remainder_ok_encodeSnappyBetterBlockAsm12B: JB three_bytes_emit_remainder_encodeSnappyBetterBlockAsm12B three_bytes_emit_remainder_encodeSnappyBetterBlockAsm12B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12B two_bytes_emit_remainder_encodeSnappyBetterBlockAsm12B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBetterBlockAsm12B JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12B one_byte_emit_remainder_encodeSnappyBetterBlockAsm12B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBetterBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -16695,73 +16713,73 @@ memmove_emit_remainder_encodeSnappyBetterBlockAsm12B: JMP emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) - JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B - -emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) + JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B + +emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm12B_memmove_move_33through64: + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm12B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm12B memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12Blarge_big_loop_back: MOVOU (SI), X4 @@ -16775,463 +16793,464 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12Blarge_big_loop JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm12Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm12B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBetterBlockAsm10B(dst []byte, src []byte) int +// func encodeSnappyBetterBlockAsm10B(dst []byte, src []byte, tmp *[20480]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBetterBlockAsm10B(SB), $20504-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x000000a0, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBetterBlockAsm10B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x000000a0, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBetterBlockAsm10B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBetterBlockAsm10B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBetterBlockAsm10B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm10B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x36, R10 - MOVL 24(SP)(R9*4), BX - MOVL 16408(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 16408(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x34, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x36, R11 + MOVL (AX)(R10*4), SI + MOVL 16384(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 16384(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm10B - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeSnappyBetterBlockAsm10B - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeSnappyBetterBlockAsm10B no_short_found_encodeSnappyBetterBlockAsm10B: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm10B - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeSnappyBetterBlockAsm10B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBetterBlockAsm10B candidateS_match_encodeSnappyBetterBlockAsm10B: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x34, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x34, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBetterBlockAsm10B - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeSnappyBetterBlockAsm10B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm10B match_extend_back_loop_encodeSnappyBetterBlockAsm10B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBetterBlockAsm10B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBetterBlockAsm10B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm10B JMP match_extend_back_loop_encodeSnappyBetterBlockAsm10B match_extend_back_end_encodeSnappyBetterBlockAsm10B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBetterBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBetterBlockAsm10B: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm10B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm10B - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm10B - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm10B - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm10B matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm10B matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm10B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm10B - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm10B - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm10B matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm10B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm10B matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm10B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm10B - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 JNE matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm10B - LEAL -4(DI), DI - LEAL 4(R11), R11 + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm10B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm10B JB match_nolit_end_encodeSnappyBetterBlockAsm10B - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm10B - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeSnappyBetterBlockAsm10B matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm10B: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeSnappyBetterBlockAsm10B - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeSnappyBetterBlockAsm10B: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeSnappyBetterBlockAsm10B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeSnappyBetterBlockAsm10B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeSnappyBetterBlockAsm10B JB three_bytes_match_emit_encodeSnappyBetterBlockAsm10B three_bytes_match_emit_encodeSnappyBetterBlockAsm10B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm10B two_bytes_match_emit_encodeSnappyBetterBlockAsm10B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeSnappyBetterBlockAsm10B JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm10B one_byte_match_emit_encodeSnappyBetterBlockAsm10B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBetterBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm10B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm10B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm10B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm10B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm10B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeSnappyBetterBlockAsm10B memmove_long_match_emit_encodeSnappyBetterBlockAsm10B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm10Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeSnappyBetterBlockAsm10B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm10B: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm10B - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm10B two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm10B: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm10B - CMPL DI, $0x00000800 + CMPL R8, $0x00000800 JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm10B - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm10B emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm10B: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm10B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm10B - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBetterBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBetterBlockAsm10B: - MOVQ $0x0000cf1bbcdcbf9b, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x34, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x36, R10 - SHLQ $0x10, R11 - IMULQ BX, R11 - SHRQ $0x34, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x36, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 16408(SP)(R10*4) - MOVL R13, 16408(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x0000cf1bbcdcbf9b, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x10, R10 + IMULQ SI, R10 + SHRQ $0x34, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x36, R11 + SHLQ $0x10, R12 + IMULQ SI, R12 + SHRQ $0x34, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x36, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 16384(AX)(R11*4) + MOVL R14, 16384(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeSnappyBetterBlockAsm10B: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeSnappyBetterBlockAsm10B - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x34, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x10, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x34, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x10, R11 + IMULQ SI, R11 + SHRQ $0x34, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeSnappyBetterBlockAsm10B emit_remainder_encodeSnappyBetterBlockAsm10B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBetterBlockAsm10B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBetterBlockAsm10B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm10B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBetterBlockAsm10B @@ -17240,26 +17259,26 @@ emit_remainder_ok_encodeSnappyBetterBlockAsm10B: JB three_bytes_emit_remainder_encodeSnappyBetterBlockAsm10B three_bytes_emit_remainder_encodeSnappyBetterBlockAsm10B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10B two_bytes_emit_remainder_encodeSnappyBetterBlockAsm10B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBetterBlockAsm10B JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10B one_byte_emit_remainder_encodeSnappyBetterBlockAsm10B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBetterBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -17275,73 +17294,73 @@ memmove_emit_remainder_encodeSnappyBetterBlockAsm10B: JMP emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm10B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm10B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm10B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm10B memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10Blarge_big_loop_back: MOVOU (SI), X4 @@ -17355,461 +17374,462 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10Blarge_big_loop JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm10Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm10B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func encodeSnappyBetterBlockAsm8B(dst []byte, src []byte) int +// func encodeSnappyBetterBlockAsm8B(dst []byte, src []byte, tmp *[5120]byte) int // Requires: BMI, SSE2 -TEXT ·encodeSnappyBetterBlockAsm8B(SB), $5144-56 - MOVQ dst_base+0(FP), AX - MOVQ $0x00000028, CX - LEAQ 24(SP), DX +TEXT ·encodeSnappyBetterBlockAsm8B(SB), $24-64 + MOVQ tmp+48(FP), AX + MOVQ dst_base+0(FP), CX + MOVQ $0x00000028, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_encodeSnappyBetterBlockAsm8B: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_encodeSnappyBetterBlockAsm8B MOVL $0x00000000, 12(SP) - MOVQ src_len+32(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX + MOVQ src_len+32(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX MOVL $0x00000000, 16(SP) - MOVQ src_base+24(FP), DX + MOVQ src_base+24(FP), BX search_loop_encodeSnappyBetterBlockAsm8B: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x04, BX - LEAL 1(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x04, SI + LEAL 1(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm8B - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ $0x9e3779b1, BX - MOVQ SI, R9 - MOVQ SI, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 - SHLQ $0x20, R10 - IMULQ BX, R10 - SHRQ $0x38, R10 - MOVL 24(SP)(R9*4), BX - MOVL 4120(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - MOVL CX, 4120(SP)(R10*4) - MOVQ (DX)(BX*1), R9 - MOVQ (DX)(DI*1), R10 - CMPQ R9, SI + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ $0x9e3779b1, SI + MOVQ DI, R10 + MOVQ DI, R11 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x36, R10 + SHLQ $0x20, R11 + IMULQ SI, R11 + SHRQ $0x38, R11 + MOVL (AX)(R10*4), SI + MOVL 4096(AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + MOVL DX, 4096(AX)(R11*4) + MOVQ (BX)(SI*1), R10 + MOVQ (BX)(R8*1), R11 + CMPQ R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm8B - CMPQ R10, SI + CMPQ R11, DI JNE no_short_found_encodeSnappyBetterBlockAsm8B - MOVL DI, BX + MOVL R8, SI JMP candidate_match_encodeSnappyBetterBlockAsm8B no_short_found_encodeSnappyBetterBlockAsm8B: - CMPL R9, SI + CMPL R10, DI JEQ candidate_match_encodeSnappyBetterBlockAsm8B - CMPL R10, SI + CMPL R11, DI JEQ candidateS_match_encodeSnappyBetterBlockAsm8B - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_encodeSnappyBetterBlockAsm8B candidateS_match_encodeSnappyBetterBlockAsm8B: - SHRQ $0x08, SI - MOVQ SI, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x36, R9 - MOVL 24(SP)(R9*4), BX - INCL CX - MOVL CX, 24(SP)(R9*4) - CMPL (DX)(BX*1), SI + SHRQ $0x08, DI + MOVQ DI, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x36, R10 + MOVL (AX)(R10*4), SI + INCL DX + MOVL DX, (AX)(R10*4) + CMPL (BX)(SI*1), DI JEQ candidate_match_encodeSnappyBetterBlockAsm8B - DECL CX - MOVL DI, BX + DECL DX + MOVL R8, SI candidate_match_encodeSnappyBetterBlockAsm8B: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm8B match_extend_back_loop_encodeSnappyBetterBlockAsm8B: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_encodeSnappyBetterBlockAsm8B - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_encodeSnappyBetterBlockAsm8B - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_encodeSnappyBetterBlockAsm8B JMP match_extend_back_loop_encodeSnappyBetterBlockAsm8B match_extend_back_end_encodeSnappyBetterBlockAsm8B: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_encodeSnappyBetterBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_dst_size_check_encodeSnappyBetterBlockAsm8B: - MOVL CX, SI - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+32(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), R9 + MOVL DX, DI + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+32(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), R10 // matchLen - XORL R11, R11 + XORL R12, R12 matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm8B: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm8B - MOVQ (R8)(R11*1), R10 - MOVQ 8(R8)(R11*1), R12 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + MOVQ 8(R9)(R12*1), R13 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm8B - XORQ 8(R9)(R11*1), R12 + XORQ 8(R10)(R12*1), R13 JNZ matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm8B - LEAL -16(DI), DI - LEAL 16(R11), R11 + LEAL -16(R8), R8 + LEAL 16(R12), R12 JMP matchlen_loopback_16_match_nolit_encodeSnappyBetterBlockAsm8B matchlen_bsf_16match_nolit_encodeSnappyBetterBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R12, R12 + TZCNTQ R13, R13 #else - BSFQ R12, R12 + BSFQ R13, R13 #endif - SARQ $0x03, R12 - LEAL 8(R11)(R12*1), R11 + SARQ $0x03, R13 + LEAL 8(R12)(R13*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm8B matchlen_match8_match_nolit_encodeSnappyBetterBlockAsm8B: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm8B - MOVQ (R8)(R11*1), R10 - XORQ (R9)(R11*1), R10 + MOVQ (R9)(R12*1), R11 + XORQ (R10)(R12*1), R11 JNZ matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm8B - LEAL -8(DI), DI - LEAL 8(R11), R11 + LEAL -8(R8), R8 + LEAL 8(R12), R12 JMP matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm8B matchlen_bsf_8_match_nolit_encodeSnappyBetterBlockAsm8B: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL (R11)(R10*1), R11 + SARQ $0x03, R11 + LEAL (R12)(R11*1), R12 JMP match_nolit_end_encodeSnappyBetterBlockAsm8B matchlen_match4_match_nolit_encodeSnappyBetterBlockAsm8B: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm8B - MOVL (R8)(R11*1), R10 - CMPL (R9)(R11*1), R10 - JNE matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm8B - LEAL -4(DI), DI - LEAL 4(R11), R11 + MOVL (R9)(R12*1), R11 + CMPL (R10)(R12*1), R11 + JNE matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm8B + LEAL -4(R8), R8 + LEAL 4(R12), R12 matchlen_match2_match_nolit_encodeSnappyBetterBlockAsm8B: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm8B JB match_nolit_end_encodeSnappyBetterBlockAsm8B - MOVW (R8)(R11*1), R10 - CMPW (R9)(R11*1), R10 + MOVW (R9)(R12*1), R11 + CMPW (R10)(R12*1), R11 JNE matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm8B - LEAL 2(R11), R11 - SUBL $0x02, DI + LEAL 2(R12), R12 + SUBL $0x02, R8 JZ match_nolit_end_encodeSnappyBetterBlockAsm8B matchlen_match1_match_nolit_encodeSnappyBetterBlockAsm8B: - MOVB (R8)(R11*1), R10 - CMPB (R9)(R11*1), R10 + MOVB (R9)(R12*1), R11 + CMPB (R10)(R12*1), R11 JNE match_nolit_end_encodeSnappyBetterBlockAsm8B - LEAL 1(R11), R11 + LEAL 1(R12), R12 match_nolit_end_encodeSnappyBetterBlockAsm8B: - MOVL CX, DI - SUBL BX, DI + MOVL DX, R8 + SUBL SI, R8 // Check if repeat - MOVL DI, 16(SP) - MOVL 12(SP), BX - CMPL BX, SI + MOVL R8, 16(SP) + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_match_emit_encodeSnappyBetterBlockAsm8B - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R9 - SUBL BX, R8 - LEAL -1(R8), BX - CMPL BX, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R10 + SUBL SI, R9 + LEAL -1(R9), SI + CMPL SI, $0x3c JB one_byte_match_emit_encodeSnappyBetterBlockAsm8B - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_match_emit_encodeSnappyBetterBlockAsm8B JB three_bytes_match_emit_encodeSnappyBetterBlockAsm8B three_bytes_match_emit_encodeSnappyBetterBlockAsm8B: - MOVB $0xf4, (AX) - MOVW BX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW SI, 1(CX) + ADDQ $0x03, CX JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm8B two_bytes_match_emit_encodeSnappyBetterBlockAsm8B: - MOVB $0xf0, (AX) - MOVB BL, 1(AX) - ADDQ $0x02, AX - CMPL BX, $0x40 + MOVB $0xf0, (CX) + MOVB SI, 1(CX) + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_match_emit_encodeSnappyBetterBlockAsm8B JMP memmove_long_match_emit_encodeSnappyBetterBlockAsm8B one_byte_match_emit_encodeSnappyBetterBlockAsm8B: - SHLB $0x02, BL - MOVB BL, (AX) - ADDQ $0x01, AX + SHLB $0x02, SI + MOVB SI, (CX) + ADDQ $0x01, CX memmove_match_emit_encodeSnappyBetterBlockAsm8B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveShort - CMPQ R8, $0x08 + CMPQ R9, $0x08 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_8 - CMPQ R8, $0x10 + CMPQ R9, $0x10 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_8through16 - CMPQ R8, $0x20 + CMPQ R9, $0x20 JBE emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_17through32 JMP emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_33through64 emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_8: - MOVQ (R9), R10 - MOVQ R10, (AX) + MOVQ (R10), R11 + MOVQ R11, (CX) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm8B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_8through16: - MOVQ (R9), R10 - MOVQ -8(R9)(R8*1), R9 - MOVQ R10, (AX) - MOVQ R9, -8(AX)(R8*1) + MOVQ (R10), R11 + MOVQ -8(R10)(R9*1), R10 + MOVQ R11, (CX) + MOVQ R10, -8(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm8B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_17through32: - MOVOU (R9), X0 - MOVOU -16(R9)(R8*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU -16(R10)(R9*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(R9*1) JMP memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm8B emit_lit_memmove_match_emit_encodeSnappyBetterBlockAsm8B_memmove_move_33through64: - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) memmove_end_copy_match_emit_encodeSnappyBetterBlockAsm8B: - MOVQ BX, AX + MOVQ SI, CX JMP emit_literal_done_match_emit_encodeSnappyBetterBlockAsm8B memmove_long_match_emit_encodeSnappyBetterBlockAsm8B: - LEAQ (AX)(R8*1), BX + LEAQ (CX)(R9*1), SI // genMemMoveLong - MOVOU (R9), X0 - MOVOU 16(R9), X1 - MOVOU -32(R9)(R8*1), X2 - MOVOU -16(R9)(R8*1), X3 - MOVQ R8, R12 - SHRQ $0x05, R12 - MOVQ AX, R10 - ANDL $0x0000001f, R10 - MOVQ $0x00000040, R13 - SUBQ R10, R13 - DECQ R12 + MOVOU (R10), X0 + MOVOU 16(R10), X1 + MOVOU -32(R10)(R9*1), X2 + MOVOU -16(R10)(R9*1), X3 + MOVQ R9, R13 + SHRQ $0x05, R13 + MOVQ CX, R11 + ANDL $0x0000001f, R11 + MOVQ $0x00000040, R14 + SUBQ R11, R14 + DECQ R13 JA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(R9)(R13*1), R10 - LEAQ -32(AX)(R13*1), R14 + LEAQ -32(R10)(R14*1), R11 + LEAQ -32(CX)(R14*1), R15 emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm8Blarge_big_loop_back: - MOVOU (R10), X4 - MOVOU 16(R10), X5 - MOVOA X4, (R14) - MOVOA X5, 16(R14) + MOVOU (R11), X4 + MOVOU 16(R11), X5 + MOVOA X4, (R15) + MOVOA X5, 16(R15) + ADDQ $0x20, R15 + ADDQ $0x20, R11 ADDQ $0x20, R14 - ADDQ $0x20, R10 - ADDQ $0x20, R13 - DECQ R12 + DECQ R13 JNA emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(R9)(R13*1), X4 - MOVOU -16(R9)(R13*1), X5 - MOVOA X4, -32(AX)(R13*1) - MOVOA X5, -16(AX)(R13*1) - ADDQ $0x20, R13 - CMPQ R8, R13 + MOVOU -32(R10)(R14*1), X4 + MOVOU -16(R10)(R14*1), X5 + MOVOA X4, -32(CX)(R14*1) + MOVOA X5, -16(CX)(R14*1) + ADDQ $0x20, R14 + CMPQ R9, R14 JAE emit_lit_memmove_long_match_emit_encodeSnappyBetterBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(R8*1) - MOVOU X3, -16(AX)(R8*1) - MOVQ BX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(R9*1) + MOVOU X3, -16(CX)(R9*1) + MOVQ SI, CX emit_literal_done_match_emit_encodeSnappyBetterBlockAsm8B: - ADDL R11, CX - ADDL $0x04, R11 - MOVL CX, 12(SP) + ADDL R12, DX + ADDL $0x04, R12 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm8B: - CMPL R11, $0x40 + CMPL R12, $0x40 JBE two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm8B - MOVB $0xee, (AX) - MOVW DI, 1(AX) - LEAL -60(R11), R11 - ADDQ $0x03, AX + MOVB $0xee, (CX) + MOVW R8, 1(CX) + LEAL -60(R12), R12 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_encodeSnappyBetterBlockAsm8B two_byte_offset_short_match_nolit_encodeSnappyBetterBlockAsm8B: - MOVL R11, BX - SHLL $0x02, BX - CMPL R11, $0x0c + MOVL R12, SI + SHLL $0x02, SI + CMPL R12, $0x0c JAE emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm8B - LEAL -15(BX), BX - MOVB DI, 1(AX) - SHRL $0x08, DI - SHLL $0x05, DI - ORL DI, BX - MOVB BL, (AX) - ADDQ $0x02, AX + LEAL -15(SI), SI + MOVB R8, 1(CX) + SHRL $0x08, R8 + SHLL $0x05, R8 + ORL R8, SI + MOVB SI, (CX) + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm8B emit_copy_three_match_nolit_encodeSnappyBetterBlockAsm8B: - LEAL -2(BX), BX - MOVB BL, (AX) - MOVW DI, 1(AX) - ADDQ $0x03, AX + LEAL -2(SI), SI + MOVB SI, (CX) + MOVW R8, 1(CX) + ADDQ $0x03, CX match_nolit_emitcopy_end_encodeSnappyBetterBlockAsm8B: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_encodeSnappyBetterBlockAsm8B - CMPQ AX, (SP) + CMPQ CX, (SP) JB match_nolit_dst_ok_encodeSnappyBetterBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET match_nolit_dst_ok_encodeSnappyBetterBlockAsm8B: - MOVQ $0x0000cf1bbcdcbf9b, BX - MOVQ $0x9e3779b1, DI - LEAQ 1(SI), SI - LEAQ -2(CX), R8 - MOVQ (DX)(SI*1), R9 - MOVQ 1(DX)(SI*1), R10 - MOVQ (DX)(R8*1), R11 - MOVQ 1(DX)(R8*1), R12 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x36, R9 - SHLQ $0x20, R10 - IMULQ DI, R10 - SHRQ $0x38, R10 - SHLQ $0x10, R11 - IMULQ BX, R11 - SHRQ $0x36, R11 - SHLQ $0x20, R12 - IMULQ DI, R12 - SHRQ $0x38, R12 - LEAQ 1(SI), DI - LEAQ 1(R8), R13 - MOVL SI, 24(SP)(R9*4) - MOVL R8, 24(SP)(R11*4) - MOVL DI, 4120(SP)(R10*4) - MOVL R13, 4120(SP)(R12*4) - LEAQ 1(R8)(SI*1), DI - SHRQ $0x01, DI - ADDQ $0x01, SI - SUBQ $0x01, R8 + MOVQ $0x0000cf1bbcdcbf9b, SI + MOVQ $0x9e3779b1, R8 + LEAQ 1(DI), DI + LEAQ -2(DX), R9 + MOVQ (BX)(DI*1), R10 + MOVQ 1(BX)(DI*1), R11 + MOVQ (BX)(R9*1), R12 + MOVQ 1(BX)(R9*1), R13 + SHLQ $0x10, R10 + IMULQ SI, R10 + SHRQ $0x36, R10 + SHLQ $0x20, R11 + IMULQ R8, R11 + SHRQ $0x38, R11 + SHLQ $0x10, R12 + IMULQ SI, R12 + SHRQ $0x36, R12 + SHLQ $0x20, R13 + IMULQ R8, R13 + SHRQ $0x38, R13 + LEAQ 1(DI), R8 + LEAQ 1(R9), R14 + MOVL DI, (AX)(R10*4) + MOVL R9, (AX)(R12*4) + MOVL R8, 4096(AX)(R11*4) + MOVL R14, 4096(AX)(R13*4) + LEAQ 1(R9)(DI*1), R8 + SHRQ $0x01, R8 + ADDQ $0x01, DI + SUBQ $0x01, R9 index_loop_encodeSnappyBetterBlockAsm8B: - CMPQ DI, R8 + CMPQ R8, R9 JAE search_loop_encodeSnappyBetterBlockAsm8B - MOVQ (DX)(SI*1), R9 - MOVQ (DX)(DI*1), R10 - SHLQ $0x10, R9 - IMULQ BX, R9 - SHRQ $0x36, R9 + MOVQ (BX)(DI*1), R10 + MOVQ (BX)(R8*1), R11 SHLQ $0x10, R10 - IMULQ BX, R10 + IMULQ SI, R10 SHRQ $0x36, R10 - MOVL SI, 24(SP)(R9*4) - MOVL DI, 24(SP)(R10*4) - ADDQ $0x02, SI + SHLQ $0x10, R11 + IMULQ SI, R11 + SHRQ $0x36, R11 + MOVL DI, (AX)(R10*4) + MOVL R8, (AX)(R11*4) ADDQ $0x02, DI + ADDQ $0x02, R8 JMP index_loop_encodeSnappyBetterBlockAsm8B emit_remainder_encodeSnappyBetterBlockAsm8B: - MOVQ src_len+32(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+32(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_encodeSnappyBetterBlockAsm8B - MOVQ $0x00000000, ret+48(FP) + MOVQ $0x00000000, ret+56(FP) RET emit_remainder_ok_encodeSnappyBetterBlockAsm8B: - MOVQ src_len+32(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+32(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm8B - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI LEAL -1(SI), DX CMPL DX, $0x3c JB one_byte_emit_remainder_encodeSnappyBetterBlockAsm8B @@ -17818,26 +17838,26 @@ emit_remainder_ok_encodeSnappyBetterBlockAsm8B: JB three_bytes_emit_remainder_encodeSnappyBetterBlockAsm8B three_bytes_emit_remainder_encodeSnappyBetterBlockAsm8B: - MOVB $0xf4, (AX) - MOVW DX, 1(AX) - ADDQ $0x03, AX + MOVB $0xf4, (CX) + MOVW DX, 1(CX) + ADDQ $0x03, CX JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8B two_bytes_emit_remainder_encodeSnappyBetterBlockAsm8B: - MOVB $0xf0, (AX) - MOVB DL, 1(AX) - ADDQ $0x02, AX + MOVB $0xf0, (CX) + MOVB DL, 1(CX) + ADDQ $0x02, CX CMPL DX, $0x40 JB memmove_emit_remainder_encodeSnappyBetterBlockAsm8B JMP memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8B one_byte_emit_remainder_encodeSnappyBetterBlockAsm8B: SHLB $0x02, DL - MOVB DL, (AX) - ADDQ $0x01, AX + MOVB DL, (CX) + ADDQ $0x01, CX memmove_emit_remainder_encodeSnappyBetterBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveShort @@ -17853,73 +17873,73 @@ memmove_emit_remainder_encodeSnappyBetterBlockAsm8B: JMP emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_33through64 emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_1or2: - MOVB (CX), SI - MOVB -1(CX)(BX*1), CL - MOVB SI, (AX) - MOVB CL, -1(AX)(BX*1) + MOVB (AX), SI + MOVB -1(AX)(BX*1), AL + MOVB SI, (CX) + MOVB AL, -1(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_3: - MOVW (CX), SI - MOVB 2(CX), CL - MOVW SI, (AX) - MOVB CL, 2(AX) + MOVW (AX), SI + MOVB 2(AX), AL + MOVW SI, (CX) + MOVB AL, 2(CX) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_4through7: - MOVL (CX), SI - MOVL -4(CX)(BX*1), CX - MOVL SI, (AX) - MOVL CX, -4(AX)(BX*1) + MOVL (AX), SI + MOVL -4(AX)(BX*1), AX + MOVL SI, (CX) + MOVL AX, -4(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_8through16: - MOVQ (CX), SI - MOVQ -8(CX)(BX*1), CX - MOVQ SI, (AX) - MOVQ CX, -8(AX)(BX*1) + MOVQ (AX), SI + MOVQ -8(AX)(BX*1), AX + MOVQ SI, (CX) + MOVQ AX, -8(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_17through32: - MOVOU (CX), X0 - MOVOU -16(CX)(BX*1), X1 - MOVOU X0, (AX) - MOVOU X1, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU -16(AX)(BX*1), X1 + MOVOU X0, (CX) + MOVOU X1, -16(CX)(BX*1) JMP memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm8B emit_lit_memmove_emit_remainder_encodeSnappyBetterBlockAsm8B_memmove_move_33through64: - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) memmove_end_copy_emit_remainder_encodeSnappyBetterBlockAsm8B: - MOVQ DX, AX + MOVQ DX, CX JMP emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm8B memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8B: - LEAQ (AX)(SI*1), DX + LEAQ (CX)(SI*1), DX MOVL SI, BX // genMemMoveLong - MOVOU (CX), X0 - MOVOU 16(CX), X1 - MOVOU -32(CX)(BX*1), X2 - MOVOU -16(CX)(BX*1), X3 + MOVOU (AX), X0 + MOVOU 16(AX), X1 + MOVOU -32(AX)(BX*1), X2 + MOVOU -16(AX)(BX*1), X3 MOVQ BX, DI SHRQ $0x05, DI - MOVQ AX, SI + MOVQ CX, SI ANDL $0x0000001f, SI MOVQ $0x00000040, R8 SUBQ SI, R8 DECQ DI JA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8Blarge_forward_sse_loop_32 - LEAQ -32(CX)(R8*1), SI - LEAQ -32(AX)(R8*1), R9 + LEAQ -32(AX)(R8*1), SI + LEAQ -32(CX)(R8*1), R9 emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8Blarge_big_loop_back: MOVOU (SI), X4 @@ -17933,1136 +17953,1142 @@ emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8Blarge_big_loop_ JNA emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8Blarge_big_loop_back emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8Blarge_forward_sse_loop_32: - MOVOU -32(CX)(R8*1), X4 - MOVOU -16(CX)(R8*1), X5 - MOVOA X4, -32(AX)(R8*1) - MOVOA X5, -16(AX)(R8*1) + MOVOU -32(AX)(R8*1), X4 + MOVOU -16(AX)(R8*1), X5 + MOVOA X4, -32(CX)(R8*1) + MOVOA X5, -16(CX)(R8*1) ADDQ $0x20, R8 CMPQ BX, R8 JAE emit_lit_memmove_long_emit_remainder_encodeSnappyBetterBlockAsm8Blarge_forward_sse_loop_32 - MOVOU X0, (AX) - MOVOU X1, 16(AX) - MOVOU X2, -32(AX)(BX*1) - MOVOU X3, -16(AX)(BX*1) - MOVQ DX, AX + MOVOU X0, (CX) + MOVOU X1, 16(CX) + MOVOU X2, -32(CX)(BX*1) + MOVOU X3, -16(CX)(BX*1) + MOVQ DX, CX emit_literal_done_emit_remainder_encodeSnappyBetterBlockAsm8B: - MOVQ dst_base+0(FP), CX - SUBQ CX, AX - MOVQ AX, ret+48(FP) + MOVQ dst_base+0(FP), AX + SUBQ AX, CX + MOVQ CX, ret+56(FP) RET -// func calcBlockSize(src []byte) int +// func calcBlockSize(src []byte, tmp *[32768]byte) int // Requires: BMI, SSE2 -TEXT ·calcBlockSize(SB), $32792-32 - XORQ AX, AX - MOVQ $0x00000100, CX - LEAQ 24(SP), DX +TEXT ·calcBlockSize(SB), $24-40 + MOVQ tmp+24(FP), AX + XORQ CX, CX + MOVQ $0x00000100, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_calcBlockSize: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_calcBlockSize MOVL $0x00000000, 12(SP) - MOVQ src_len+8(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+0(FP), DX + MOVQ src_len+8(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+0(FP), BX search_loop_calcBlockSize: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x05, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x05, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_calcBlockSize - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x33, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x10, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x33, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x10, R9 - IMULQ R8, R9 - SHRQ $0x33, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x10, R11 + IMULQ R9, R11 + SHRQ $0x33, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x10, R10 + IMULQ R9, R10 + SHRQ $0x33, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_calcBlockSize - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_calcBlockSize repeat_extend_back_loop_calcBlockSize: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_calcBlockSize - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_calcBlockSize - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_calcBlockSize repeat_extend_back_end_calcBlockSize: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 5(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 5(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_calcBlockSize - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET repeat_dst_size_check_calcBlockSize: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_calcBlockSize - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_calcBlockSize - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_calcBlockSize - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB three_bytes_repeat_emit_calcBlockSize - CMPL BX, $0x01000000 + CMPL SI, $0x01000000 JB four_bytes_repeat_emit_calcBlockSize - ADDQ $0x05, AX + ADDQ $0x05, CX JMP memmove_long_repeat_emit_calcBlockSize four_bytes_repeat_emit_calcBlockSize: - ADDQ $0x04, AX + ADDQ $0x04, CX JMP memmove_long_repeat_emit_calcBlockSize three_bytes_repeat_emit_calcBlockSize: - ADDQ $0x03, AX + ADDQ $0x03, CX JMP memmove_long_repeat_emit_calcBlockSize two_bytes_repeat_emit_calcBlockSize: - ADDQ $0x02, AX - CMPL BX, $0x40 + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_calcBlockSize JMP memmove_long_repeat_emit_calcBlockSize one_byte_repeat_emit_calcBlockSize: - ADDQ $0x01, AX + ADDQ $0x01, CX memmove_repeat_emit_calcBlockSize: - LEAQ (AX)(DI*1), AX + LEAQ (CX)(R8*1), CX JMP emit_literal_done_repeat_emit_calcBlockSize memmove_long_repeat_emit_calcBlockSize: - LEAQ (AX)(DI*1), AX + LEAQ (CX)(R8*1), CX emit_literal_done_repeat_emit_calcBlockSize: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+8(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+8(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_calcBlockSize: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_calcBlockSize - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_calcBlockSize - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_calcBlockSize - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_calcBlockSize matchlen_bsf_16repeat_extend_calcBlockSize: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_calcBlockSize matchlen_match8_repeat_extend_calcBlockSize: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_calcBlockSize - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_calcBlockSize - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_calcBlockSize matchlen_bsf_8_repeat_extend_calcBlockSize: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_calcBlockSize matchlen_match4_repeat_extend_calcBlockSize: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_calcBlockSize - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_calcBlockSize - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_calcBlockSize: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_calcBlockSize JB repeat_extend_forward_end_calcBlockSize - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_calcBlockSize - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_calcBlockSize matchlen_match1_repeat_extend_calcBlockSize: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_calcBlockSize - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_calcBlockSize: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy - CMPL SI, $0x00010000 + CMPL DI, $0x00010000 JB two_byte_offset_repeat_as_copy_calcBlockSize four_bytes_loop_back_repeat_as_copy_calcBlockSize: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE four_bytes_remain_repeat_as_copy_calcBlockSize - LEAL -64(BX), BX - ADDQ $0x05, AX - CMPL BX, $0x04 + LEAL -64(SI), SI + ADDQ $0x05, CX + CMPL SI, $0x04 JB four_bytes_remain_repeat_as_copy_calcBlockSize JMP four_bytes_loop_back_repeat_as_copy_calcBlockSize four_bytes_remain_repeat_as_copy_calcBlockSize: - TESTL BX, BX + TESTL SI, SI JZ repeat_end_emit_calcBlockSize - XORL BX, BX - ADDQ $0x05, AX + XORL SI, SI + ADDQ $0x05, CX JMP repeat_end_emit_calcBlockSize two_byte_offset_repeat_as_copy_calcBlockSize: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_calcBlockSize - LEAL -60(BX), BX - ADDQ $0x03, AX + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_calcBlockSize two_byte_offset_short_repeat_as_copy_calcBlockSize: - MOVL BX, DI - SHLL $0x02, DI - CMPL BX, $0x0c + MOVL SI, R8 + SHLL $0x02, R8 + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_calcBlockSize - CMPL SI, $0x00000800 + CMPL DI, $0x00000800 JAE emit_copy_three_repeat_as_copy_calcBlockSize - ADDQ $0x02, AX + ADDQ $0x02, CX JMP repeat_end_emit_calcBlockSize emit_copy_three_repeat_as_copy_calcBlockSize: - ADDQ $0x03, AX + ADDQ $0x03, CX repeat_end_emit_calcBlockSize: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_calcBlockSize no_repeat_found_calcBlockSize: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_calcBlockSize - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_calcBlockSize - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_calcBlockSize - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_calcBlockSize candidate3_match_calcBlockSize: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_calcBlockSize candidate2_match_calcBlockSize: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_calcBlockSize: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_calcBlockSize match_extend_back_loop_calcBlockSize: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_calcBlockSize - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_calcBlockSize - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_calcBlockSize JMP match_extend_back_loop_calcBlockSize match_extend_back_end_calcBlockSize: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 5(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 5(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_calcBlockSize - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET match_dst_size_check_calcBlockSize: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_calcBlockSize - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), SI - CMPL SI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), DI + CMPL DI, $0x3c JB one_byte_match_emit_calcBlockSize - CMPL SI, $0x00000100 + CMPL DI, $0x00000100 JB two_bytes_match_emit_calcBlockSize - CMPL SI, $0x00010000 + CMPL DI, $0x00010000 JB three_bytes_match_emit_calcBlockSize - CMPL SI, $0x01000000 + CMPL DI, $0x01000000 JB four_bytes_match_emit_calcBlockSize - ADDQ $0x05, AX + ADDQ $0x05, CX JMP memmove_long_match_emit_calcBlockSize four_bytes_match_emit_calcBlockSize: - ADDQ $0x04, AX + ADDQ $0x04, CX JMP memmove_long_match_emit_calcBlockSize three_bytes_match_emit_calcBlockSize: - ADDQ $0x03, AX + ADDQ $0x03, CX JMP memmove_long_match_emit_calcBlockSize two_bytes_match_emit_calcBlockSize: - ADDQ $0x02, AX - CMPL SI, $0x40 + ADDQ $0x02, CX + CMPL DI, $0x40 JB memmove_match_emit_calcBlockSize JMP memmove_long_match_emit_calcBlockSize one_byte_match_emit_calcBlockSize: - ADDQ $0x01, AX + ADDQ $0x01, CX memmove_match_emit_calcBlockSize: - LEAQ (AX)(R8*1), AX + LEAQ (CX)(R9*1), CX JMP emit_literal_done_match_emit_calcBlockSize memmove_long_match_emit_calcBlockSize: - LEAQ (AX)(R8*1), AX + LEAQ (CX)(R9*1), CX emit_literal_done_match_emit_calcBlockSize: match_nolit_loop_calcBlockSize: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+8(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+8(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_calcBlockSize: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_calcBlockSize - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_calcBlockSize - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_calcBlockSize - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_calcBlockSize matchlen_bsf_16match_nolit_calcBlockSize: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_calcBlockSize matchlen_match8_match_nolit_calcBlockSize: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_calcBlockSize - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_calcBlockSize - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_calcBlockSize matchlen_bsf_8_match_nolit_calcBlockSize: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_calcBlockSize matchlen_match4_match_nolit_calcBlockSize: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_calcBlockSize - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_calcBlockSize - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_calcBlockSize: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_calcBlockSize JB match_nolit_end_calcBlockSize - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_calcBlockSize - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_calcBlockSize matchlen_match1_match_nolit_calcBlockSize: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_calcBlockSize - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_calcBlockSize: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy - CMPL BX, $0x00010000 + CMPL SI, $0x00010000 JB two_byte_offset_match_nolit_calcBlockSize four_bytes_loop_back_match_nolit_calcBlockSize: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE four_bytes_remain_match_nolit_calcBlockSize - LEAL -64(R9), R9 - ADDQ $0x05, AX - CMPL R9, $0x04 + LEAL -64(R10), R10 + ADDQ $0x05, CX + CMPL R10, $0x04 JB four_bytes_remain_match_nolit_calcBlockSize JMP four_bytes_loop_back_match_nolit_calcBlockSize four_bytes_remain_match_nolit_calcBlockSize: - TESTL R9, R9 + TESTL R10, R10 JZ match_nolit_emitcopy_end_calcBlockSize - XORL BX, BX - ADDQ $0x05, AX + XORL SI, SI + ADDQ $0x05, CX JMP match_nolit_emitcopy_end_calcBlockSize two_byte_offset_match_nolit_calcBlockSize: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_calcBlockSize - LEAL -60(R9), R9 - ADDQ $0x03, AX + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_calcBlockSize two_byte_offset_short_match_nolit_calcBlockSize: - MOVL R9, SI - SHLL $0x02, SI - CMPL R9, $0x0c + MOVL R10, DI + SHLL $0x02, DI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_calcBlockSize - CMPL BX, $0x00000800 + CMPL SI, $0x00000800 JAE emit_copy_three_match_nolit_calcBlockSize - ADDQ $0x02, AX + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_calcBlockSize emit_copy_three_match_nolit_calcBlockSize: - ADDQ $0x03, AX + ADDQ $0x03, CX match_nolit_emitcopy_end_calcBlockSize: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_calcBlockSize - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_calcBlockSize - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET match_nolit_dst_ok_calcBlockSize: - MOVQ $0x0000cf1bbcdcbf9b, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x10, DI - IMULQ R8, DI - SHRQ $0x33, DI - SHLQ $0x10, BX - IMULQ R8, BX - SHRQ $0x33, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x0000cf1bbcdcbf9b, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x10, R8 + IMULQ R9, R8 + SHRQ $0x33, R8 + SHLQ $0x10, SI + IMULQ R9, SI + SHRQ $0x33, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_calcBlockSize - INCL CX + INCL DX JMP search_loop_calcBlockSize emit_remainder_calcBlockSize: - MOVQ src_len+8(FP), CX - SUBL 12(SP), CX - LEAQ 5(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+8(FP), AX + SUBL 12(SP), AX + LEAQ 5(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_calcBlockSize - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET emit_remainder_ok_calcBlockSize: - MOVQ src_len+8(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+8(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_calcBlockSize - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI - LEAL -1(SI), CX - CMPL CX, $0x3c + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI + LEAL -1(SI), AX + CMPL AX, $0x3c JB one_byte_emit_remainder_calcBlockSize - CMPL CX, $0x00000100 + CMPL AX, $0x00000100 JB two_bytes_emit_remainder_calcBlockSize - CMPL CX, $0x00010000 + CMPL AX, $0x00010000 JB three_bytes_emit_remainder_calcBlockSize - CMPL CX, $0x01000000 + CMPL AX, $0x01000000 JB four_bytes_emit_remainder_calcBlockSize - ADDQ $0x05, AX + ADDQ $0x05, CX JMP memmove_long_emit_remainder_calcBlockSize four_bytes_emit_remainder_calcBlockSize: - ADDQ $0x04, AX + ADDQ $0x04, CX JMP memmove_long_emit_remainder_calcBlockSize three_bytes_emit_remainder_calcBlockSize: - ADDQ $0x03, AX + ADDQ $0x03, CX JMP memmove_long_emit_remainder_calcBlockSize two_bytes_emit_remainder_calcBlockSize: - ADDQ $0x02, AX - CMPL CX, $0x40 + ADDQ $0x02, CX + CMPL AX, $0x40 JB memmove_emit_remainder_calcBlockSize JMP memmove_long_emit_remainder_calcBlockSize one_byte_emit_remainder_calcBlockSize: - ADDQ $0x01, AX + ADDQ $0x01, CX memmove_emit_remainder_calcBlockSize: - LEAQ (AX)(SI*1), AX + LEAQ (CX)(SI*1), AX + MOVQ AX, CX JMP emit_literal_done_emit_remainder_calcBlockSize memmove_long_emit_remainder_calcBlockSize: - LEAQ (AX)(SI*1), AX + LEAQ (CX)(SI*1), AX + MOVQ AX, CX emit_literal_done_emit_remainder_calcBlockSize: - MOVQ AX, ret+24(FP) + MOVQ CX, ret+32(FP) RET -// func calcBlockSizeSmall(src []byte) int +// func calcBlockSizeSmall(src []byte, tmp *[2048]byte) int // Requires: BMI, SSE2 -TEXT ·calcBlockSizeSmall(SB), $2072-32 - XORQ AX, AX - MOVQ $0x00000010, CX - LEAQ 24(SP), DX +TEXT ·calcBlockSizeSmall(SB), $24-40 + MOVQ tmp+24(FP), AX + XORQ CX, CX + MOVQ $0x00000010, DX + MOVQ AX, BX PXOR X0, X0 zero_loop_calcBlockSizeSmall: - MOVOU X0, (DX) - MOVOU X0, 16(DX) - MOVOU X0, 32(DX) - MOVOU X0, 48(DX) - MOVOU X0, 64(DX) - MOVOU X0, 80(DX) - MOVOU X0, 96(DX) - MOVOU X0, 112(DX) - ADDQ $0x80, DX - DECQ CX + MOVOU X0, (BX) + MOVOU X0, 16(BX) + MOVOU X0, 32(BX) + MOVOU X0, 48(BX) + MOVOU X0, 64(BX) + MOVOU X0, 80(BX) + MOVOU X0, 96(BX) + MOVOU X0, 112(BX) + ADDQ $0x80, BX + DECQ DX JNZ zero_loop_calcBlockSizeSmall MOVL $0x00000000, 12(SP) - MOVQ src_len+8(FP), CX - LEAQ -9(CX), DX - LEAQ -8(CX), BX - MOVL BX, 8(SP) - SHRQ $0x05, CX - SUBL CX, DX - LEAQ (AX)(DX*1), DX - MOVQ DX, (SP) - MOVL $0x00000001, CX - MOVL CX, 16(SP) - MOVQ src_base+0(FP), DX + MOVQ src_len+8(FP), DX + LEAQ -9(DX), BX + LEAQ -8(DX), SI + MOVL SI, 8(SP) + SHRQ $0x05, DX + SUBL DX, BX + LEAQ (CX)(BX*1), BX + MOVQ BX, (SP) + MOVL $0x00000001, DX + MOVL DX, 16(SP) + MOVQ src_base+0(FP), BX search_loop_calcBlockSizeSmall: - MOVL CX, BX - SUBL 12(SP), BX - SHRL $0x04, BX - LEAL 4(CX)(BX*1), BX - CMPL BX, 8(SP) + MOVL DX, SI + SUBL 12(SP), SI + SHRL $0x04, SI + LEAL 4(DX)(SI*1), SI + CMPL SI, 8(SP) JAE emit_remainder_calcBlockSizeSmall - MOVQ (DX)(CX*1), SI - MOVL BX, 20(SP) - MOVQ $0x9e3779b1, R8 - MOVQ SI, R9 - MOVQ SI, R10 - SHRQ $0x08, R10 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x37, R9 + MOVQ (BX)(DX*1), DI + MOVL SI, 20(SP) + MOVQ $0x9e3779b1, R9 + MOVQ DI, R10 + MOVQ DI, R11 + SHRQ $0x08, R11 SHLQ $0x20, R10 - IMULQ R8, R10 + IMULQ R9, R10 SHRQ $0x37, R10 - MOVL 24(SP)(R9*4), BX - MOVL 24(SP)(R10*4), DI - MOVL CX, 24(SP)(R9*4) - LEAL 1(CX), R9 - MOVL R9, 24(SP)(R10*4) - MOVQ SI, R9 - SHRQ $0x10, R9 - SHLQ $0x20, R9 - IMULQ R8, R9 - SHRQ $0x37, R9 - MOVL CX, R8 - SUBL 16(SP), R8 - MOVL 1(DX)(R8*1), R10 - MOVQ SI, R8 - SHRQ $0x08, R8 - CMPL R8, R10 + SHLQ $0x20, R11 + IMULQ R9, R11 + SHRQ $0x37, R11 + MOVL (AX)(R10*4), SI + MOVL (AX)(R11*4), R8 + MOVL DX, (AX)(R10*4) + LEAL 1(DX), R10 + MOVL R10, (AX)(R11*4) + MOVQ DI, R10 + SHRQ $0x10, R10 + SHLQ $0x20, R10 + IMULQ R9, R10 + SHRQ $0x37, R10 + MOVL DX, R9 + SUBL 16(SP), R9 + MOVL 1(BX)(R9*1), R11 + MOVQ DI, R9 + SHRQ $0x08, R9 + CMPL R9, R11 JNE no_repeat_found_calcBlockSizeSmall - LEAL 1(CX), SI - MOVL 12(SP), BX - MOVL SI, DI - SUBL 16(SP), DI + LEAL 1(DX), DI + MOVL 12(SP), SI + MOVL DI, R8 + SUBL 16(SP), R8 JZ repeat_extend_back_end_calcBlockSizeSmall repeat_extend_back_loop_calcBlockSizeSmall: - CMPL SI, BX + CMPL DI, SI JBE repeat_extend_back_end_calcBlockSizeSmall - MOVB -1(DX)(DI*1), R8 - MOVB -1(DX)(SI*1), R9 - CMPB R8, R9 + MOVB -1(BX)(R8*1), R9 + MOVB -1(BX)(DI*1), R10 + CMPB R9, R10 JNE repeat_extend_back_end_calcBlockSizeSmall - LEAL -1(SI), SI - DECL DI + LEAL -1(DI), DI + DECL R8 JNZ repeat_extend_back_loop_calcBlockSizeSmall repeat_extend_back_end_calcBlockSizeSmall: - MOVL SI, BX - SUBL 12(SP), BX - LEAQ 3(AX)(BX*1), BX - CMPQ BX, (SP) + MOVL DI, SI + SUBL 12(SP), SI + LEAQ 3(CX)(SI*1), SI + CMPQ SI, (SP) JB repeat_dst_size_check_calcBlockSizeSmall - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET repeat_dst_size_check_calcBlockSizeSmall: - MOVL 12(SP), BX - CMPL BX, SI + MOVL 12(SP), SI + CMPL SI, DI JEQ emit_literal_done_repeat_emit_calcBlockSizeSmall - MOVL SI, DI - MOVL SI, 12(SP) - LEAQ (DX)(BX*1), R8 - SUBL BX, DI - LEAL -1(DI), BX - CMPL BX, $0x3c + MOVL DI, R8 + MOVL DI, 12(SP) + LEAQ (BX)(SI*1), R9 + SUBL SI, R8 + LEAL -1(R8), SI + CMPL SI, $0x3c JB one_byte_repeat_emit_calcBlockSizeSmall - CMPL BX, $0x00000100 + CMPL SI, $0x00000100 JB two_bytes_repeat_emit_calcBlockSizeSmall JB three_bytes_repeat_emit_calcBlockSizeSmall three_bytes_repeat_emit_calcBlockSizeSmall: - ADDQ $0x03, AX + ADDQ $0x03, CX JMP memmove_long_repeat_emit_calcBlockSizeSmall two_bytes_repeat_emit_calcBlockSizeSmall: - ADDQ $0x02, AX - CMPL BX, $0x40 + ADDQ $0x02, CX + CMPL SI, $0x40 JB memmove_repeat_emit_calcBlockSizeSmall JMP memmove_long_repeat_emit_calcBlockSizeSmall one_byte_repeat_emit_calcBlockSizeSmall: - ADDQ $0x01, AX + ADDQ $0x01, CX memmove_repeat_emit_calcBlockSizeSmall: - LEAQ (AX)(DI*1), AX + LEAQ (CX)(R8*1), CX JMP emit_literal_done_repeat_emit_calcBlockSizeSmall memmove_long_repeat_emit_calcBlockSizeSmall: - LEAQ (AX)(DI*1), AX + LEAQ (CX)(R8*1), CX emit_literal_done_repeat_emit_calcBlockSizeSmall: - ADDL $0x05, CX - MOVL CX, BX - SUBL 16(SP), BX - MOVQ src_len+8(FP), DI - SUBL CX, DI - LEAQ (DX)(CX*1), R8 - LEAQ (DX)(BX*1), BX + ADDL $0x05, DX + MOVL DX, SI + SUBL 16(SP), SI + MOVQ src_len+8(FP), R8 + SUBL DX, R8 + LEAQ (BX)(DX*1), R9 + LEAQ (BX)(SI*1), SI // matchLen - XORL R10, R10 + XORL R11, R11 matchlen_loopback_16_repeat_extend_calcBlockSizeSmall: - CMPL DI, $0x10 + CMPL R8, $0x10 JB matchlen_match8_repeat_extend_calcBlockSizeSmall - MOVQ (R8)(R10*1), R9 - MOVQ 8(R8)(R10*1), R11 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + MOVQ 8(R9)(R11*1), R12 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_calcBlockSizeSmall - XORQ 8(BX)(R10*1), R11 + XORQ 8(SI)(R11*1), R12 JNZ matchlen_bsf_16repeat_extend_calcBlockSizeSmall - LEAL -16(DI), DI - LEAL 16(R10), R10 + LEAL -16(R8), R8 + LEAL 16(R11), R11 JMP matchlen_loopback_16_repeat_extend_calcBlockSizeSmall matchlen_bsf_16repeat_extend_calcBlockSizeSmall: #ifdef GOAMD64_v3 - TZCNTQ R11, R11 + TZCNTQ R12, R12 #else - BSFQ R11, R11 + BSFQ R12, R12 #endif - SARQ $0x03, R11 - LEAL 8(R10)(R11*1), R10 + SARQ $0x03, R12 + LEAL 8(R11)(R12*1), R11 JMP repeat_extend_forward_end_calcBlockSizeSmall matchlen_match8_repeat_extend_calcBlockSizeSmall: - CMPL DI, $0x08 + CMPL R8, $0x08 JB matchlen_match4_repeat_extend_calcBlockSizeSmall - MOVQ (R8)(R10*1), R9 - XORQ (BX)(R10*1), R9 + MOVQ (R9)(R11*1), R10 + XORQ (SI)(R11*1), R10 JNZ matchlen_bsf_8_repeat_extend_calcBlockSizeSmall - LEAL -8(DI), DI - LEAL 8(R10), R10 + LEAL -8(R8), R8 + LEAL 8(R11), R11 JMP matchlen_match4_repeat_extend_calcBlockSizeSmall matchlen_bsf_8_repeat_extend_calcBlockSizeSmall: #ifdef GOAMD64_v3 - TZCNTQ R9, R9 + TZCNTQ R10, R10 #else - BSFQ R9, R9 + BSFQ R10, R10 #endif - SARQ $0x03, R9 - LEAL (R10)(R9*1), R10 + SARQ $0x03, R10 + LEAL (R11)(R10*1), R11 JMP repeat_extend_forward_end_calcBlockSizeSmall matchlen_match4_repeat_extend_calcBlockSizeSmall: - CMPL DI, $0x04 + CMPL R8, $0x04 JB matchlen_match2_repeat_extend_calcBlockSizeSmall - MOVL (R8)(R10*1), R9 - CMPL (BX)(R10*1), R9 + MOVL (R9)(R11*1), R10 + CMPL (SI)(R11*1), R10 JNE matchlen_match2_repeat_extend_calcBlockSizeSmall - LEAL -4(DI), DI - LEAL 4(R10), R10 + LEAL -4(R8), R8 + LEAL 4(R11), R11 matchlen_match2_repeat_extend_calcBlockSizeSmall: - CMPL DI, $0x01 + CMPL R8, $0x01 JE matchlen_match1_repeat_extend_calcBlockSizeSmall JB repeat_extend_forward_end_calcBlockSizeSmall - MOVW (R8)(R10*1), R9 - CMPW (BX)(R10*1), R9 + MOVW (R9)(R11*1), R10 + CMPW (SI)(R11*1), R10 JNE matchlen_match1_repeat_extend_calcBlockSizeSmall - LEAL 2(R10), R10 - SUBL $0x02, DI + LEAL 2(R11), R11 + SUBL $0x02, R8 JZ repeat_extend_forward_end_calcBlockSizeSmall matchlen_match1_repeat_extend_calcBlockSizeSmall: - MOVB (R8)(R10*1), R9 - CMPB (BX)(R10*1), R9 + MOVB (R9)(R11*1), R10 + CMPB (SI)(R11*1), R10 JNE repeat_extend_forward_end_calcBlockSizeSmall - LEAL 1(R10), R10 + LEAL 1(R11), R11 repeat_extend_forward_end_calcBlockSizeSmall: - ADDL R10, CX - MOVL CX, BX - SUBL SI, BX - MOVL 16(SP), SI + ADDL R11, DX + MOVL DX, SI + SUBL DI, SI + MOVL 16(SP), DI // emitCopy two_byte_offset_repeat_as_copy_calcBlockSizeSmall: - CMPL BX, $0x40 + CMPL SI, $0x40 JBE two_byte_offset_short_repeat_as_copy_calcBlockSizeSmall - LEAL -60(BX), BX - ADDQ $0x03, AX + LEAL -60(SI), SI + ADDQ $0x03, CX JMP two_byte_offset_repeat_as_copy_calcBlockSizeSmall two_byte_offset_short_repeat_as_copy_calcBlockSizeSmall: - MOVL BX, SI - SHLL $0x02, SI - CMPL BX, $0x0c + MOVL SI, DI + SHLL $0x02, DI + CMPL SI, $0x0c JAE emit_copy_three_repeat_as_copy_calcBlockSizeSmall - ADDQ $0x02, AX + ADDQ $0x02, CX JMP repeat_end_emit_calcBlockSizeSmall emit_copy_three_repeat_as_copy_calcBlockSizeSmall: - ADDQ $0x03, AX + ADDQ $0x03, CX repeat_end_emit_calcBlockSizeSmall: - MOVL CX, 12(SP) + MOVL DX, 12(SP) JMP search_loop_calcBlockSizeSmall no_repeat_found_calcBlockSizeSmall: - CMPL (DX)(BX*1), SI + CMPL (BX)(SI*1), DI JEQ candidate_match_calcBlockSizeSmall - SHRQ $0x08, SI - MOVL 24(SP)(R9*4), BX - LEAL 2(CX), R8 - CMPL (DX)(DI*1), SI + SHRQ $0x08, DI + MOVL (AX)(R10*4), SI + LEAL 2(DX), R9 + CMPL (BX)(R8*1), DI JEQ candidate2_match_calcBlockSizeSmall - MOVL R8, 24(SP)(R9*4) - SHRQ $0x08, SI - CMPL (DX)(BX*1), SI + MOVL R9, (AX)(R10*4) + SHRQ $0x08, DI + CMPL (BX)(SI*1), DI JEQ candidate3_match_calcBlockSizeSmall - MOVL 20(SP), CX + MOVL 20(SP), DX JMP search_loop_calcBlockSizeSmall candidate3_match_calcBlockSizeSmall: - ADDL $0x02, CX + ADDL $0x02, DX JMP candidate_match_calcBlockSizeSmall candidate2_match_calcBlockSizeSmall: - MOVL R8, 24(SP)(R9*4) - INCL CX - MOVL DI, BX + MOVL R9, (AX)(R10*4) + INCL DX + MOVL R8, SI candidate_match_calcBlockSizeSmall: - MOVL 12(SP), SI - TESTL BX, BX + MOVL 12(SP), DI + TESTL SI, SI JZ match_extend_back_end_calcBlockSizeSmall match_extend_back_loop_calcBlockSizeSmall: - CMPL CX, SI + CMPL DX, DI JBE match_extend_back_end_calcBlockSizeSmall - MOVB -1(DX)(BX*1), DI - MOVB -1(DX)(CX*1), R8 - CMPB DI, R8 + MOVB -1(BX)(SI*1), R8 + MOVB -1(BX)(DX*1), R9 + CMPB R8, R9 JNE match_extend_back_end_calcBlockSizeSmall - LEAL -1(CX), CX - DECL BX + LEAL -1(DX), DX + DECL SI JZ match_extend_back_end_calcBlockSizeSmall JMP match_extend_back_loop_calcBlockSizeSmall match_extend_back_end_calcBlockSizeSmall: - MOVL CX, SI - SUBL 12(SP), SI - LEAQ 3(AX)(SI*1), SI - CMPQ SI, (SP) + MOVL DX, DI + SUBL 12(SP), DI + LEAQ 3(CX)(DI*1), DI + CMPQ DI, (SP) JB match_dst_size_check_calcBlockSizeSmall - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET match_dst_size_check_calcBlockSizeSmall: - MOVL CX, SI - MOVL 12(SP), DI - CMPL DI, SI + MOVL DX, DI + MOVL 12(SP), R8 + CMPL R8, DI JEQ emit_literal_done_match_emit_calcBlockSizeSmall - MOVL SI, R8 - MOVL SI, 12(SP) - LEAQ (DX)(DI*1), SI - SUBL DI, R8 - LEAL -1(R8), SI - CMPL SI, $0x3c + MOVL DI, R9 + MOVL DI, 12(SP) + LEAQ (BX)(R8*1), DI + SUBL R8, R9 + LEAL -1(R9), DI + CMPL DI, $0x3c JB one_byte_match_emit_calcBlockSizeSmall - CMPL SI, $0x00000100 + CMPL DI, $0x00000100 JB two_bytes_match_emit_calcBlockSizeSmall JB three_bytes_match_emit_calcBlockSizeSmall three_bytes_match_emit_calcBlockSizeSmall: - ADDQ $0x03, AX + ADDQ $0x03, CX JMP memmove_long_match_emit_calcBlockSizeSmall two_bytes_match_emit_calcBlockSizeSmall: - ADDQ $0x02, AX - CMPL SI, $0x40 + ADDQ $0x02, CX + CMPL DI, $0x40 JB memmove_match_emit_calcBlockSizeSmall JMP memmove_long_match_emit_calcBlockSizeSmall one_byte_match_emit_calcBlockSizeSmall: - ADDQ $0x01, AX + ADDQ $0x01, CX memmove_match_emit_calcBlockSizeSmall: - LEAQ (AX)(R8*1), AX + LEAQ (CX)(R9*1), CX JMP emit_literal_done_match_emit_calcBlockSizeSmall memmove_long_match_emit_calcBlockSizeSmall: - LEAQ (AX)(R8*1), AX + LEAQ (CX)(R9*1), CX emit_literal_done_match_emit_calcBlockSizeSmall: match_nolit_loop_calcBlockSizeSmall: - MOVL CX, SI - SUBL BX, SI - MOVL SI, 16(SP) - ADDL $0x04, CX - ADDL $0x04, BX - MOVQ src_len+8(FP), SI - SUBL CX, SI - LEAQ (DX)(CX*1), DI - LEAQ (DX)(BX*1), BX + MOVL DX, DI + SUBL SI, DI + MOVL DI, 16(SP) + ADDL $0x04, DX + ADDL $0x04, SI + MOVQ src_len+8(FP), DI + SUBL DX, DI + LEAQ (BX)(DX*1), R8 + LEAQ (BX)(SI*1), SI // matchLen - XORL R9, R9 + XORL R10, R10 matchlen_loopback_16_match_nolit_calcBlockSizeSmall: - CMPL SI, $0x10 + CMPL DI, $0x10 JB matchlen_match8_match_nolit_calcBlockSizeSmall - MOVQ (DI)(R9*1), R8 - MOVQ 8(DI)(R9*1), R10 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + MOVQ 8(R8)(R10*1), R11 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_calcBlockSizeSmall - XORQ 8(BX)(R9*1), R10 + XORQ 8(SI)(R10*1), R11 JNZ matchlen_bsf_16match_nolit_calcBlockSizeSmall - LEAL -16(SI), SI - LEAL 16(R9), R9 + LEAL -16(DI), DI + LEAL 16(R10), R10 JMP matchlen_loopback_16_match_nolit_calcBlockSizeSmall matchlen_bsf_16match_nolit_calcBlockSizeSmall: #ifdef GOAMD64_v3 - TZCNTQ R10, R10 + TZCNTQ R11, R11 #else - BSFQ R10, R10 + BSFQ R11, R11 #endif - SARQ $0x03, R10 - LEAL 8(R9)(R10*1), R9 + SARQ $0x03, R11 + LEAL 8(R10)(R11*1), R10 JMP match_nolit_end_calcBlockSizeSmall matchlen_match8_match_nolit_calcBlockSizeSmall: - CMPL SI, $0x08 + CMPL DI, $0x08 JB matchlen_match4_match_nolit_calcBlockSizeSmall - MOVQ (DI)(R9*1), R8 - XORQ (BX)(R9*1), R8 + MOVQ (R8)(R10*1), R9 + XORQ (SI)(R10*1), R9 JNZ matchlen_bsf_8_match_nolit_calcBlockSizeSmall - LEAL -8(SI), SI - LEAL 8(R9), R9 + LEAL -8(DI), DI + LEAL 8(R10), R10 JMP matchlen_match4_match_nolit_calcBlockSizeSmall matchlen_bsf_8_match_nolit_calcBlockSizeSmall: #ifdef GOAMD64_v3 - TZCNTQ R8, R8 + TZCNTQ R9, R9 #else - BSFQ R8, R8 + BSFQ R9, R9 #endif - SARQ $0x03, R8 - LEAL (R9)(R8*1), R9 + SARQ $0x03, R9 + LEAL (R10)(R9*1), R10 JMP match_nolit_end_calcBlockSizeSmall matchlen_match4_match_nolit_calcBlockSizeSmall: - CMPL SI, $0x04 + CMPL DI, $0x04 JB matchlen_match2_match_nolit_calcBlockSizeSmall - MOVL (DI)(R9*1), R8 - CMPL (BX)(R9*1), R8 + MOVL (R8)(R10*1), R9 + CMPL (SI)(R10*1), R9 JNE matchlen_match2_match_nolit_calcBlockSizeSmall - LEAL -4(SI), SI - LEAL 4(R9), R9 + LEAL -4(DI), DI + LEAL 4(R10), R10 matchlen_match2_match_nolit_calcBlockSizeSmall: - CMPL SI, $0x01 + CMPL DI, $0x01 JE matchlen_match1_match_nolit_calcBlockSizeSmall JB match_nolit_end_calcBlockSizeSmall - MOVW (DI)(R9*1), R8 - CMPW (BX)(R9*1), R8 + MOVW (R8)(R10*1), R9 + CMPW (SI)(R10*1), R9 JNE matchlen_match1_match_nolit_calcBlockSizeSmall - LEAL 2(R9), R9 - SUBL $0x02, SI + LEAL 2(R10), R10 + SUBL $0x02, DI JZ match_nolit_end_calcBlockSizeSmall matchlen_match1_match_nolit_calcBlockSizeSmall: - MOVB (DI)(R9*1), R8 - CMPB (BX)(R9*1), R8 + MOVB (R8)(R10*1), R9 + CMPB (SI)(R10*1), R9 JNE match_nolit_end_calcBlockSizeSmall - LEAL 1(R9), R9 + LEAL 1(R10), R10 match_nolit_end_calcBlockSizeSmall: - ADDL R9, CX - MOVL 16(SP), BX - ADDL $0x04, R9 - MOVL CX, 12(SP) + ADDL R10, DX + MOVL 16(SP), SI + ADDL $0x04, R10 + MOVL DX, 12(SP) // emitCopy two_byte_offset_match_nolit_calcBlockSizeSmall: - CMPL R9, $0x40 + CMPL R10, $0x40 JBE two_byte_offset_short_match_nolit_calcBlockSizeSmall - LEAL -60(R9), R9 - ADDQ $0x03, AX + LEAL -60(R10), R10 + ADDQ $0x03, CX JMP two_byte_offset_match_nolit_calcBlockSizeSmall two_byte_offset_short_match_nolit_calcBlockSizeSmall: - MOVL R9, BX - SHLL $0x02, BX - CMPL R9, $0x0c + MOVL R10, SI + SHLL $0x02, SI + CMPL R10, $0x0c JAE emit_copy_three_match_nolit_calcBlockSizeSmall - ADDQ $0x02, AX + ADDQ $0x02, CX JMP match_nolit_emitcopy_end_calcBlockSizeSmall emit_copy_three_match_nolit_calcBlockSizeSmall: - ADDQ $0x03, AX + ADDQ $0x03, CX match_nolit_emitcopy_end_calcBlockSizeSmall: - CMPL CX, 8(SP) + CMPL DX, 8(SP) JAE emit_remainder_calcBlockSizeSmall - MOVQ -2(DX)(CX*1), SI - CMPQ AX, (SP) + MOVQ -2(BX)(DX*1), DI + CMPQ CX, (SP) JB match_nolit_dst_ok_calcBlockSizeSmall - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET match_nolit_dst_ok_calcBlockSizeSmall: - MOVQ $0x9e3779b1, R8 - MOVQ SI, DI - SHRQ $0x10, SI - MOVQ SI, BX - SHLQ $0x20, DI - IMULQ R8, DI - SHRQ $0x37, DI - SHLQ $0x20, BX - IMULQ R8, BX - SHRQ $0x37, BX - LEAL -2(CX), R8 - LEAQ 24(SP)(BX*4), R9 - MOVL (R9), BX - MOVL R8, 24(SP)(DI*4) - MOVL CX, (R9) - CMPL (DX)(BX*1), SI + MOVQ $0x9e3779b1, R9 + MOVQ DI, R8 + SHRQ $0x10, DI + MOVQ DI, SI + SHLQ $0x20, R8 + IMULQ R9, R8 + SHRQ $0x37, R8 + SHLQ $0x20, SI + IMULQ R9, SI + SHRQ $0x37, SI + LEAL -2(DX), R9 + LEAQ (AX)(SI*4), R10 + MOVL (R10), SI + MOVL R9, (AX)(R8*4) + MOVL DX, (R10) + CMPL (BX)(SI*1), DI JEQ match_nolit_loop_calcBlockSizeSmall - INCL CX + INCL DX JMP search_loop_calcBlockSizeSmall emit_remainder_calcBlockSizeSmall: - MOVQ src_len+8(FP), CX - SUBL 12(SP), CX - LEAQ 3(AX)(CX*1), CX - CMPQ CX, (SP) + MOVQ src_len+8(FP), AX + SUBL 12(SP), AX + LEAQ 3(CX)(AX*1), AX + CMPQ AX, (SP) JB emit_remainder_ok_calcBlockSizeSmall - MOVQ $0x00000000, ret+24(FP) + MOVQ $0x00000000, ret+32(FP) RET emit_remainder_ok_calcBlockSizeSmall: - MOVQ src_len+8(FP), CX - MOVL 12(SP), BX - CMPL BX, CX + MOVQ src_len+8(FP), AX + MOVL 12(SP), DX + CMPL DX, AX JEQ emit_literal_done_emit_remainder_calcBlockSizeSmall - MOVL CX, SI - MOVL CX, 12(SP) - LEAQ (DX)(BX*1), CX - SUBL BX, SI - LEAL -1(SI), CX - CMPL CX, $0x3c + MOVL AX, SI + MOVL AX, 12(SP) + LEAQ (BX)(DX*1), AX + SUBL DX, SI + LEAL -1(SI), AX + CMPL AX, $0x3c JB one_byte_emit_remainder_calcBlockSizeSmall - CMPL CX, $0x00000100 + CMPL AX, $0x00000100 JB two_bytes_emit_remainder_calcBlockSizeSmall JB three_bytes_emit_remainder_calcBlockSizeSmall three_bytes_emit_remainder_calcBlockSizeSmall: - ADDQ $0x03, AX + ADDQ $0x03, CX JMP memmove_long_emit_remainder_calcBlockSizeSmall two_bytes_emit_remainder_calcBlockSizeSmall: - ADDQ $0x02, AX - CMPL CX, $0x40 + ADDQ $0x02, CX + CMPL AX, $0x40 JB memmove_emit_remainder_calcBlockSizeSmall JMP memmove_long_emit_remainder_calcBlockSizeSmall one_byte_emit_remainder_calcBlockSizeSmall: - ADDQ $0x01, AX + ADDQ $0x01, CX memmove_emit_remainder_calcBlockSizeSmall: - LEAQ (AX)(SI*1), AX + LEAQ (CX)(SI*1), AX + MOVQ AX, CX JMP emit_literal_done_emit_remainder_calcBlockSizeSmall memmove_long_emit_remainder_calcBlockSizeSmall: - LEAQ (AX)(SI*1), AX + LEAQ (CX)(SI*1), AX + MOVQ AX, CX emit_literal_done_emit_remainder_calcBlockSizeSmall: - MOVQ AX, ret+24(FP) + MOVQ CX, ret+32(FP) RET // func emitLiteral(dst []byte, lit []byte) int @@ -19783,7 +19809,7 @@ TEXT ·cvtLZ4BlockAsm(SB), NOSPLIT, $0-64 MOVQ src_base+24(FP), DX MOVQ src_len+32(FP), BX LEAQ (DX)(BX*1), BX - LEAQ -10(AX)(CX*1), CX + LEAQ -8(AX)(CX*1), CX XORQ DI, DI lz4_s2_loop: @@ -20266,7 +20292,7 @@ TEXT ·cvtLZ4sBlockAsm(SB), NOSPLIT, $0-64 MOVQ src_base+24(FP), DX MOVQ src_len+32(FP), BX LEAQ (DX)(BX*1), BX - LEAQ -10(AX)(CX*1), CX + LEAQ -8(AX)(CX*1), CX XORQ DI, DI lz4s_s2_loop: @@ -20751,7 +20777,7 @@ TEXT ·cvtLZ4BlockSnappyAsm(SB), NOSPLIT, $0-64 MOVQ src_base+24(FP), DX MOVQ src_len+32(FP), BX LEAQ (DX)(BX*1), BX - LEAQ -10(AX)(CX*1), CX + LEAQ -8(AX)(CX*1), CX lz4_snappy_loop: CMPQ DX, BX @@ -21017,7 +21043,7 @@ TEXT ·cvtLZ4sBlockSnappyAsm(SB), NOSPLIT, $0-64 MOVQ src_base+24(FP), DX MOVQ src_len+32(FP), BX LEAQ (DX)(BX*1), BX - LEAQ -10(AX)(CX*1), CX + LEAQ -8(AX)(CX*1), CX lz4s_snappy_loop: CMPQ DX, BX diff --git a/vendor/github.com/klauspost/compress/zstd/encoder.go b/vendor/github.com/klauspost/compress/zstd/encoder.go index a79c4a52..8f8223cd 100644 --- a/vendor/github.com/klauspost/compress/zstd/encoder.go +++ b/vendor/github.com/klauspost/compress/zstd/encoder.go @@ -6,6 +6,7 @@ package zstd import ( "crypto/rand" + "errors" "fmt" "io" "math" @@ -149,6 +150,9 @@ func (e *Encoder) ResetContentSize(w io.Writer, size int64) { // and write CRC if requested. func (e *Encoder) Write(p []byte) (n int, err error) { s := &e.state + if s.eofWritten { + return 0, ErrEncoderClosed + } for len(p) > 0 { if len(p)+len(s.filling) < e.o.blockSize { if e.o.crc { @@ -288,6 +292,9 @@ func (e *Encoder) nextBlock(final bool) error { s.filling, s.current, s.previous = s.previous[:0], s.filling, s.current s.nInput += int64(len(s.current)) s.wg.Add(1) + if final { + s.eofWritten = true + } go func(src []byte) { if debugEncoder { println("Adding block,", len(src), "bytes, final:", final) @@ -303,9 +310,6 @@ func (e *Encoder) nextBlock(final bool) error { blk := enc.Block() enc.Encode(blk, src) blk.last = final - if final { - s.eofWritten = true - } // Wait for pending writes. s.wWg.Wait() if s.writeErr != nil { @@ -401,12 +405,20 @@ func (e *Encoder) Flush() error { if len(s.filling) > 0 { err := e.nextBlock(false) if err != nil { + // Ignore Flush after Close. + if errors.Is(s.err, ErrEncoderClosed) { + return nil + } return err } } s.wg.Wait() s.wWg.Wait() if s.err != nil { + // Ignore Flush after Close. + if errors.Is(s.err, ErrEncoderClosed) { + return nil + } return s.err } return s.writeErr @@ -422,6 +434,9 @@ func (e *Encoder) Close() error { } err := e.nextBlock(true) if err != nil { + if errors.Is(s.err, ErrEncoderClosed) { + return nil + } return err } if s.frameContentSize > 0 { @@ -459,6 +474,11 @@ func (e *Encoder) Close() error { } _, s.err = s.w.Write(frame) } + if s.err == nil { + s.err = ErrEncoderClosed + return nil + } + return s.err } diff --git a/vendor/github.com/klauspost/compress/zstd/zstd.go b/vendor/github.com/klauspost/compress/zstd/zstd.go index 4be7cc73..066bef2a 100644 --- a/vendor/github.com/klauspost/compress/zstd/zstd.go +++ b/vendor/github.com/klauspost/compress/zstd/zstd.go @@ -88,6 +88,10 @@ var ( // Close has been called. ErrDecoderClosed = errors.New("decoder used after Close") + // ErrEncoderClosed will be returned if the Encoder was used after + // Close has been called. + ErrEncoderClosed = errors.New("encoder used after Close") + // ErrDecoderNilInput is returned when a nil Reader was provided // and an operation other than Reset/DecodeAll/Close was attempted. ErrDecoderNilInput = errors.New("nil input provided as reader") diff --git a/vendor/github.com/minio/minio-go/v7/Makefile b/vendor/github.com/minio/minio-go/v7/Makefile index 68444aa6..9e4ddc4c 100644 --- a/vendor/github.com/minio/minio-go/v7/Makefile +++ b/vendor/github.com/minio/minio-go/v7/Makefile @@ -32,6 +32,10 @@ functional-test: @GO111MODULE=on go build -race functional_tests.go @SERVER_ENDPOINT=localhost:9000 ACCESS_KEY=minioadmin SECRET_KEY=minioadmin ENABLE_HTTPS=1 MINT_MODE=full ./functional_tests +functional-test-notls: + @GO111MODULE=on go build -race functional_tests.go + @SERVER_ENDPOINT=localhost:9000 ACCESS_KEY=minioadmin SECRET_KEY=minioadmin ENABLE_HTTPS=0 MINT_MODE=full ./functional_tests + clean: @echo "Cleaning up all the generated files" @find . -name '*.test' | xargs rm -fv diff --git a/vendor/github.com/minio/minio-go/v7/README.md b/vendor/github.com/minio/minio-go/v7/README.md index 82f70a13..be7963c5 100644 --- a/vendor/github.com/minio/minio-go/v7/README.md +++ b/vendor/github.com/minio/minio-go/v7/README.md @@ -253,7 +253,7 @@ The full API Reference is available here. * [setbucketencryption.go](https://github.com/minio/minio-go/blob/master/examples/s3/setbucketencryption.go) * [getbucketencryption.go](https://github.com/minio/minio-go/blob/master/examples/s3/getbucketencryption.go) -* [deletebucketencryption.go](https://github.com/minio/minio-go/blob/master/examples/s3/deletebucketencryption.go) +* [removebucketencryption.go](https://github.com/minio/minio-go/blob/master/examples/s3/removebucketencryption.go) ### Full Examples : Bucket replication Operations diff --git a/vendor/github.com/minio/minio-go/v7/api-put-object-streaming.go b/vendor/github.com/minio/minio-go/v7/api-put-object-streaming.go index eef976c8..dac4c0ef 100644 --- a/vendor/github.com/minio/minio-go/v7/api-put-object-streaming.go +++ b/vendor/github.com/minio/minio-go/v7/api-put-object-streaming.go @@ -52,7 +52,7 @@ func (c *Client) putObjectMultipartStream(ctx context.Context, bucketName, objec } else { info, err = c.putObjectMultipartStreamOptionalChecksum(ctx, bucketName, objectName, reader, size, opts) } - if err != nil { + if err != nil && s3utils.IsGoogleEndpoint(*c.endpointURL) { errResp := ToErrorResponse(err) // Verify if multipart functionality is not available, if not // fall back to single PutObject operation. diff --git a/vendor/github.com/minio/minio-go/v7/api-put-object.go b/vendor/github.com/minio/minio-go/v7/api-put-object.go index d769648a..10131a5b 100644 --- a/vendor/github.com/minio/minio-go/v7/api-put-object.go +++ b/vendor/github.com/minio/minio-go/v7/api-put-object.go @@ -45,6 +45,8 @@ const ( ReplicationStatusFailed ReplicationStatus = "FAILED" // ReplicationStatusReplica indicates object is a replica of a source ReplicationStatusReplica ReplicationStatus = "REPLICA" + // ReplicationStatusReplicaEdge indicates object is a replica of a edge source + ReplicationStatusReplicaEdge ReplicationStatus = "REPLICA-EDGE" ) // Empty returns true if no replication status set. diff --git a/vendor/github.com/minio/minio-go/v7/api.go b/vendor/github.com/minio/minio-go/v7/api.go index 1d6b6650..380ec4fd 100644 --- a/vendor/github.com/minio/minio-go/v7/api.go +++ b/vendor/github.com/minio/minio-go/v7/api.go @@ -99,6 +99,7 @@ type Client struct { healthStatus int32 trailingHeaderSupport bool + maxRetries int } // Options for New method @@ -123,12 +124,16 @@ type Options struct { // Custom hash routines. Leave nil to use standard. CustomMD5 func() md5simd.Hasher CustomSHA256 func() md5simd.Hasher + + // Number of times a request is retried. Defaults to 10 retries if this option is not configured. + // Set to 1 to disable retries. + MaxRetries int } // Global constants. const ( libraryName = "minio-go" - libraryVersion = "v7.0.77" + libraryVersion = "v7.0.80" ) // User Agent should always following the below style. @@ -278,6 +283,11 @@ func privateNew(endpoint string, opts *Options) (*Client, error) { // healthcheck is not initialized clnt.healthStatus = unknown + clnt.maxRetries = MaxRetry + if opts.MaxRetries > 0 { + clnt.maxRetries = opts.MaxRetries + } + // Return. return clnt, nil } @@ -590,9 +600,9 @@ func (c *Client) executeMethod(ctx context.Context, method string, metadata requ return nil, errors.New(c.endpointURL.String() + " is offline.") } - var retryable bool // Indicates if request can be retried. - var bodySeeker io.Seeker // Extracted seeker from io.Reader. - reqRetry := MaxRetry // Indicates how many times we can retry the request + var retryable bool // Indicates if request can be retried. + var bodySeeker io.Seeker // Extracted seeker from io.Reader. + var reqRetry = c.maxRetries // Indicates how many times we can retry the request if metadata.contentBody != nil { // Check if body is seekable then it is retryable. diff --git a/vendor/github.com/minio/minio-go/v7/functional_tests.go b/vendor/github.com/minio/minio-go/v7/functional_tests.go index 780dc899..c0180b36 100644 --- a/vendor/github.com/minio/minio-go/v7/functional_tests.go +++ b/vendor/github.com/minio/minio-go/v7/functional_tests.go @@ -3565,16 +3565,10 @@ func validateObjectAttributeRequest(OA *minio.ObjectAttributes, opts *minio.Obje } } - hasFullObjectChecksum := true - if OA.Checksum.ChecksumCRC32 == "" { - if OA.Checksum.ChecksumCRC32C == "" { - if OA.Checksum.ChecksumSHA1 == "" { - if OA.Checksum.ChecksumSHA256 == "" { - hasFullObjectChecksum = false - } - } - } - } + hasFullObjectChecksum := (OA.Checksum.ChecksumCRC32 != "" || + OA.Checksum.ChecksumCRC32C != "" || + OA.Checksum.ChecksumSHA1 != "" || + OA.Checksum.ChecksumSHA256 != "") if test.HasFullChecksum { if !hasFullObjectChecksum { diff --git a/vendor/github.com/minio/minio-go/v7/pkg/credentials/sts_web_identity.go b/vendor/github.com/minio/minio-go/v7/pkg/credentials/sts_web_identity.go index 596d9515..f1c76c78 100644 --- a/vendor/github.com/minio/minio-go/v7/pkg/credentials/sts_web_identity.go +++ b/vendor/github.com/minio/minio-go/v7/pkg/credentials/sts_web_identity.go @@ -25,6 +25,7 @@ import ( "io" "net/http" "net/url" + "os" "strconv" "strings" "time" @@ -85,29 +86,59 @@ type STSWebIdentity struct { // assuming. RoleARN string + // Policy is the policy where the credentials should be limited too. + Policy string + // roleSessionName is the identifier for the assumed role session. roleSessionName string } // NewSTSWebIdentity returns a pointer to a new // Credentials object wrapping the STSWebIdentity. -func NewSTSWebIdentity(stsEndpoint string, getWebIDTokenExpiry func() (*WebIdentityToken, error)) (*Credentials, error) { +func NewSTSWebIdentity(stsEndpoint string, getWebIDTokenExpiry func() (*WebIdentityToken, error), opts ...func(*STSWebIdentity)) (*Credentials, error) { if stsEndpoint == "" { return nil, errors.New("STS endpoint cannot be empty") } if getWebIDTokenExpiry == nil { return nil, errors.New("Web ID token and expiry retrieval function should be defined") } - return New(&STSWebIdentity{ + i := &STSWebIdentity{ Client: &http.Client{ Transport: http.DefaultTransport, }, STSEndpoint: stsEndpoint, GetWebIDTokenExpiry: getWebIDTokenExpiry, - }), nil + } + for _, o := range opts { + o(i) + } + return New(i), nil +} + +// NewKubernetesIdentity returns a pointer to a new +// Credentials object using the Kubernetes service account +func NewKubernetesIdentity(stsEndpoint string, opts ...func(*STSWebIdentity)) (*Credentials, error) { + return NewSTSWebIdentity(stsEndpoint, func() (*WebIdentityToken, error) { + token, err := os.ReadFile("/var/run/secrets/kubernetes.io/serviceaccount/token") + if err != nil { + return nil, err + } + + return &WebIdentityToken{ + Token: string(token), + }, nil + }, opts...) +} + +// WithPolicy option will enforce that the returned credentials +// will be scoped down to the specified policy +func WithPolicy(policy string) func(*STSWebIdentity) { + return func(i *STSWebIdentity) { + i.Policy = policy + } } -func getWebIdentityCredentials(clnt *http.Client, endpoint, roleARN, roleSessionName string, +func getWebIdentityCredentials(clnt *http.Client, endpoint, roleARN, roleSessionName string, policy string, getWebIDTokenExpiry func() (*WebIdentityToken, error), ) (AssumeRoleWithWebIdentityResponse, error) { idToken, err := getWebIDTokenExpiry() @@ -133,6 +164,9 @@ func getWebIdentityCredentials(clnt *http.Client, endpoint, roleARN, roleSession if idToken.Expiry > 0 { v.Set("DurationSeconds", fmt.Sprintf("%d", idToken.Expiry)) } + if policy != "" { + v.Set("Policy", policy) + } v.Set("Version", STSVersion) u, err := url.Parse(endpoint) @@ -183,7 +217,7 @@ func getWebIdentityCredentials(clnt *http.Client, endpoint, roleARN, roleSession // Retrieve retrieves credentials from the MinIO service. // Error will be returned if the request fails. func (m *STSWebIdentity) Retrieve() (Value, error) { - a, err := getWebIdentityCredentials(m.Client, m.STSEndpoint, m.RoleARN, m.roleSessionName, m.GetWebIDTokenExpiry) + a, err := getWebIdentityCredentials(m.Client, m.STSEndpoint, m.RoleARN, m.roleSessionName, m.Policy, m.GetWebIDTokenExpiry) if err != nil { return Value{}, err } diff --git a/vendor/github.com/minio/minio-go/v7/pkg/lifecycle/lifecycle.go b/vendor/github.com/minio/minio-go/v7/pkg/lifecycle/lifecycle.go index e706b57d..344af2b7 100644 --- a/vendor/github.com/minio/minio-go/v7/pkg/lifecycle/lifecycle.go +++ b/vendor/github.com/minio/minio-go/v7/pkg/lifecycle/lifecycle.go @@ -434,12 +434,34 @@ func (de DelMarkerExpiration) MarshalXML(enc *xml.Encoder, start xml.StartElemen return enc.EncodeElement(delMarkerExp(de), start) } +// AllVersionsExpiration represents AllVersionsExpiration actions element in an ILM policy +type AllVersionsExpiration struct { + XMLName xml.Name `xml:"AllVersionsExpiration" json:"-"` + Days int `xml:"Days,omitempty" json:"Days,omitempty"` + DeleteMarker ExpireDeleteMarker `xml:"DeleteMarker,omitempty" json:"DeleteMarker,omitempty"` +} + +// IsNull returns true if days field is 0 +func (e AllVersionsExpiration) IsNull() bool { + return e.Days == 0 +} + +// MarshalXML satisfies xml.Marshaler to provide custom encoding +func (e AllVersionsExpiration) MarshalXML(enc *xml.Encoder, start xml.StartElement) error { + if e.IsNull() { + return nil + } + type allVersionsExp AllVersionsExpiration + return enc.EncodeElement(allVersionsExp(e), start) +} + // MarshalJSON customizes json encoding by omitting empty values func (r Rule) MarshalJSON() ([]byte, error) { type rule struct { AbortIncompleteMultipartUpload *AbortIncompleteMultipartUpload `json:"AbortIncompleteMultipartUpload,omitempty"` Expiration *Expiration `json:"Expiration,omitempty"` DelMarkerExpiration *DelMarkerExpiration `json:"DelMarkerExpiration,omitempty"` + AllVersionsExpiration *AllVersionsExpiration `json:"AllVersionsExpiration,omitempty"` ID string `json:"ID"` RuleFilter *Filter `json:"Filter,omitempty"` NoncurrentVersionExpiration *NoncurrentVersionExpiration `json:"NoncurrentVersionExpiration,omitempty"` @@ -475,6 +497,9 @@ func (r Rule) MarshalJSON() ([]byte, error) { if !r.NoncurrentVersionTransition.isNull() { newr.NoncurrentVersionTransition = &r.NoncurrentVersionTransition } + if !r.AllVersionsExpiration.IsNull() { + newr.AllVersionsExpiration = &r.AllVersionsExpiration + } return json.Marshal(newr) } @@ -485,6 +510,7 @@ type Rule struct { AbortIncompleteMultipartUpload AbortIncompleteMultipartUpload `xml:"AbortIncompleteMultipartUpload,omitempty" json:"AbortIncompleteMultipartUpload,omitempty"` Expiration Expiration `xml:"Expiration,omitempty" json:"Expiration,omitempty"` DelMarkerExpiration DelMarkerExpiration `xml:"DelMarkerExpiration,omitempty" json:"DelMarkerExpiration,omitempty"` + AllVersionsExpiration AllVersionsExpiration `xml:"AllVersionsExpiration,omitempty" json:"AllVersionsExpiration,omitempty"` ID string `xml:"ID" json:"ID"` RuleFilter Filter `xml:"Filter,omitempty" json:"Filter,omitempty"` NoncurrentVersionExpiration NoncurrentVersionExpiration `xml:"NoncurrentVersionExpiration,omitempty" json:"NoncurrentVersionExpiration,omitempty"` diff --git a/vendor/github.com/minio/minio-go/v7/pkg/tags/tags.go b/vendor/github.com/minio/minio-go/v7/pkg/tags/tags.go index 7a84a6f3..33465c63 100644 --- a/vendor/github.com/minio/minio-go/v7/pkg/tags/tags.go +++ b/vendor/github.com/minio/minio-go/v7/pkg/tags/tags.go @@ -69,7 +69,7 @@ const ( // https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/Using_Tags.html#tag-restrictions // borrowed from this article and also testing various ASCII characters following regex // is supported by AWS S3 for both tags and values. -var validTagKeyValue = regexp.MustCompile(`^[a-zA-Z0-9-+\-._:/@ ]+$`) +var validTagKeyValue = regexp.MustCompile(`^[a-zA-Z0-9-+\-._:/@ =]+$`) func checkKey(key string) error { if len(key) == 0 { diff --git a/vendor/github.com/swaggo/swag/Dockerfile b/vendor/github.com/swaggo/swag/Dockerfile index 410fe31e..5ea91343 100644 --- a/vendor/github.com/swaggo/swag/Dockerfile +++ b/vendor/github.com/swaggo/swag/Dockerfile @@ -1,7 +1,7 @@ # Dockerfile References: https://docs.docker.com/engine/reference/builder/ # Start from the latest golang base image -FROM golang:1.20-alpine as builder +FROM --platform=$BUILDPLATFORM golang:1.21-alpine as builder # Set the Current Working Directory inside the container WORKDIR /app @@ -15,12 +15,18 @@ RUN go mod download # Copy the source from the current directory to the Working Directory inside the container COPY . . +# Configure go compiler target platform +ARG TARGETOS +ARG TARGETARCH +ENV GOARCH=$TARGETARCH \ + GOOS=$TARGETOS + # Build the Go app RUN CGO_ENABLED=0 GOOS=linux go build -v -a -installsuffix cgo -o swag cmd/swag/main.go ######## Start a new stage from scratch ####### -FROM scratch +FROM --platform=$TARGETPLATFORM scratch WORKDIR /code/ diff --git a/vendor/github.com/swaggo/swag/Makefile b/vendor/github.com/swaggo/swag/Makefile index 0d8175da..85dc3623 100644 --- a/vendor/github.com/swaggo/swag/Makefile +++ b/vendor/github.com/swaggo/swag/Makefile @@ -17,8 +17,6 @@ BINARY_NAME:=swag PACKAGES:=$(shell $(GOLIST) github.com/swaggo/swag github.com/swaggo/swag/cmd/swag github.com/swaggo/swag/gen github.com/swaggo/swag/format) GOFILES:=$(shell find . -name "*.go" -type f) -export GO111MODULE := on - all: test build .PHONY: build @@ -57,17 +55,8 @@ clean: deps: $(GOMODTIDY) -.PHONY: devel-deps -devel-deps: - GO111MODULE=off $(GOGET) -v -u \ - golang.org/x/lint/golint - -.PHONY: lint -lint: devel-deps - for PKG in $(PACKAGES); do golint -set_exit_status $$PKG || exit 1; done; - .PHONY: vet -vet: deps devel-deps +vet: deps $(GOVET) $(PACKAGES) .PHONY: fmt diff --git a/vendor/github.com/swaggo/swag/README.md b/vendor/github.com/swaggo/swag/README.md index 054e36ec..7504f6e7 100644 --- a/vendor/github.com/swaggo/swag/README.md +++ b/vendor/github.com/swaggo/swag/README.md @@ -31,7 +31,8 @@ Swag converts Go annotations to Swagger Documentation 2.0. We've created a varie - [User defined structure with an array type](#user-defined-structure-with-an-array-type) - [Function scoped struct declaration](#function-scoped-struct-declaration) - [Model composition in response](#model-composition-in-response) - - [Add a headers in response](#add-a-headers-in-response) + - [Add request headers](#add-request-headers) + - [Add response headers](#add-response-headers) - [Use multiple path params](#use-multiple-path-params) - [Example value of struct](#example-value-of-struct) - [SchemaExample of body](#schemaexample-of-body) @@ -55,7 +56,7 @@ Swag converts Go annotations to Swagger Documentation 2.0. We've created a varie ```sh go install github.com/swaggo/swag/cmd/swag@latest ``` -To build from source you need [Go](https://golang.org/dl/) (1.18 or newer). +To build from source you need [Go](https://golang.org/dl/) (1.19 or newer). Alternatively you can run the docker image: ```sh @@ -103,6 +104,7 @@ OPTIONS: --outputTypes value, --ot value Output types of generated files (docs.go, swagger.json, swagger.yaml) like go,json,yaml (default: "go,json,yaml") --parseVendor Parse go files in 'vendor' folder, disabled by default (default: false) --parseDependency, --pd Parse go files inside dependency folder, disabled by default (default: false) + --parseDependencyLevel, --pdl Enhancement of '--parseDependency', parse go files inside dependency folder, 0 disabled, 1 only parse models, 2 only parse operations, 3 parse all (default: 0) --markdownFiles value, --md value Parse folder containing markdown files to use as description, disabled by default --codeExampleFiles value, --cef value Parse folder containing code example files to use for the x-codeSamples extension, disabled by default --parseInternal Parse go files in internal packages, disabled by default (default: false) @@ -113,9 +115,10 @@ OPTIONS: --overridesFile value File to read global type overrides from. (default: ".swaggo") --parseGoList Parse dependency via 'go list' (default: true) --tags value, -t value A comma-separated list of tags to filter the APIs for which the documentation is generated.Special case if the tag is prefixed with the '!' character then the APIs with that tag will be excluded - --templateDelims value, --td value Provide custom delimeters for Go template generation. The format is leftDelim,rightDelim. For example: "[[,]]" + --templateDelims value, --td value Provide custom delimiters for Go template generation. The format is leftDelim,rightDelim. For example: "[[,]]" --collectionFormat value, --cf value Set default collection format (default: "csv") --state value Initial state for the state machine (default: ""), @HostState in root file, @State in other files + --parseFuncBody Parse API info within body of functions in go files, disabled by default (default: false) --help, -h show help (default: false) ``` @@ -645,7 +648,14 @@ type DeepObject struct { //in `proto` package } @success 200 {object} jsonresult.JSONResult{data1=proto.Order{data=proto.DeepObject},data2=[]proto.Order{data=[]proto.DeepObject}} "desc" ``` -### Add a headers in response +### Add response request + +```go +// @Param X-MyHeader header string true "MyHeader must be set for valid response" +// @Param X-API-VERSION header string true "API version eg.: 1.0" +``` + +### Add response headers ```go // @Success 200 {string} string "ok" @@ -931,7 +941,7 @@ func GetPosts(w http.ResponseWriter, r *http.Request) { _ = web.GenericNestedResponse[types.Post]{} } ``` -See [this file](https://github.com/swaggo/swag/blob/master/testdata/generics_nested/api/api.go) for more details +See [this file](https://github.com/swaggo/swag/blob/master/testdata/generics_nested/api/api.go) for more details and other examples. ### Change the default Go Template action delimiters @@ -946,6 +956,17 @@ swag init -g http/api.go -td "[[,]]" ``` The new delimiter is a string with the format "``,``". +### Parse Internal and Dependency Packages + +If the struct is defined in a dependency package, use `--parseDependency`. + +If the struct is defined in your main project, use `--parseInternal`. + +if you want to include both internal and from dependencies use both flags +``` +swag init --parseDependency --parseInternal +``` + ## About the Project This project was inspired by [yvasiyarov/swagger](https://github.com/yvasiyarov/swagger) but we simplified the usage and added support a variety of [web frameworks](#supported-web-frameworks). Gopher image source is [tenntenn/gopher-stickers](https://github.com/tenntenn/gopher-stickers). It has licenses [creative commons licensing](http://creativecommons.org/licenses/by/3.0/deed.en). ## Contributors diff --git a/vendor/github.com/swaggo/swag/README_pt.md b/vendor/github.com/swaggo/swag/README_pt.md index 7f95066b..060b2c97 100644 --- a/vendor/github.com/swaggo/swag/README_pt.md +++ b/vendor/github.com/swaggo/swag/README_pt.md @@ -54,7 +54,7 @@ Swag converte anotações Go para Documentação Swagger 2.0. Criámos uma varie ```sh go install github.com/swaggo/swag/cmd/swag@latest ``` -Para construir a partir da fonte é necessário [Go](https://golang.org/dl/) (1.18 ou mais recente). +Para construir a partir da fonte é necessário [Go](https://golang.org/dl/) (1.19 ou mais recente). Ou descarregar um binário pré-compilado a partir da [página de lançamento](https://github.com/swaggo/swag/releases). diff --git a/vendor/github.com/swaggo/swag/README_zh-CN.md b/vendor/github.com/swaggo/swag/README_zh-CN.md index 87d600b5..24a219c9 100644 --- a/vendor/github.com/swaggo/swag/README_zh-CN.md +++ b/vendor/github.com/swaggo/swag/README_zh-CN.md @@ -50,7 +50,7 @@ Swag将Go的注释转换为Swagger2.0文档。我们为流行的 [Go Web Framewo go install github.com/swaggo/swag/cmd/swag@latest ``` -从源码开始构建的话，需要有Go环境（1.18及以上版本）。 +从源码开始构建的话，需要有Go环境（1.19及以上版本）。 或者从github的release页面下载预编译好的二进制文件。 @@ -90,6 +90,7 @@ OPTIONS: --output value, -o value 文件(swagger.json, swagger.yaml and doc.go)输出目录 (默认: "./docs") --parseVendor 是否解析vendor目录里的go源文件，默认不 --parseDependency 是否解析依赖目录中的go源文件，默认不 + --parseDependencyLevel, --pdl 对'--parseDependency'参数进行增强, 是否解析依赖目录中的go源文件, 0 不解析, 1 只解析对象模型, 2 只解析API, 3 对象模型和API都解析 (default: 0) --markdownFiles value, --md value 指定API的描述信息所使用的markdown文件所在的目录 --generatedTime 是否输出时间到输出文件docs.go的顶部，默认是 --codeExampleFiles value, --cef value 解析包含用于 x-codeSamples 扩展的代码示例文件的文件夹，默认禁用 diff --git a/vendor/github.com/swaggo/swag/field_parser.go b/vendor/github.com/swaggo/swag/field_parser.go index 9b24e787..de8945de 100644 --- a/vendor/github.com/swaggo/swag/field_parser.go +++ b/vendor/github.com/swaggo/swag/field_parser.go @@ -65,35 +65,38 @@ func (ps *tagBaseFieldParser) ShouldSkip() bool { return false } -func (ps *tagBaseFieldParser) FieldName() (string, error) { - var name string +func (ps *tagBaseFieldParser) FieldNames() ([]string, error) { + if len(ps.field.Names) <= 1 { + // if embedded but with a json/form name ?? + if ps.field.Tag != nil { + // json:"tag,hoge" + name := strings.TrimSpace(strings.Split(ps.tag.Get(jsonTag), ",")[0]) + if name != "" { + return []string{name}, nil + } - if ps.field.Tag != nil { - // json:"tag,hoge" - name = strings.TrimSpace(strings.Split(ps.tag.Get(jsonTag), ",")[0]) - if name != "" { - return name, nil + // use "form" tag over json tag + name = ps.FormName() + if name != "" { + return []string{name}, nil + } } - - // use "form" tag over json tag - name = ps.FormName() - if name != "" { - return name, nil + if len(ps.field.Names) == 0 { + return nil, nil } } - - if ps.field.Names == nil { - return "", nil - } - - switch ps.p.PropNamingStrategy { - case SnakeCase: - return toSnakeCase(ps.field.Names[0].Name), nil - case PascalCase: - return ps.field.Names[0].Name, nil - default: - return toLowerCamelCase(ps.field.Names[0].Name), nil + var names = make([]string, 0, len(ps.field.Names)) + for _, name := range ps.field.Names { + switch ps.p.PropNamingStrategy { + case SnakeCase: + names = append(names, toSnakeCase(name.Name)) + case PascalCase: + names = append(names, name.Name) + default: + names = append(names, toLowerCamelCase(name.Name)) + } } + return names, nil } func (ps *tagBaseFieldParser) firstTagValue(tag string) string { @@ -169,6 +172,7 @@ func (ps *tagBaseFieldParser) CustomSchema() (*spec.Schema, error) { } type structField struct { + title string schemaType string arrayType string formatType string @@ -274,6 +278,7 @@ func (ps *tagBaseFieldParser) complementSchema(schema *spec.Schema, types []stri field := &structField{ schemaType: types[0], formatType: ps.tag.Get(formatTag), + title: ps.tag.Get(titleTag), } if len(types) > 1 && (types[0] == ARRAY || types[0] == OBJECT) { @@ -414,6 +419,7 @@ func (ps *tagBaseFieldParser) complementSchema(schema *spec.Schema, types []stri if field.schemaType != ARRAY { schema.Format = field.formatType } + schema.Title = field.title extensionsTagValue := ps.tag.Get(extensionsTag) if extensionsTagValue != "" { diff --git a/vendor/github.com/swaggo/swag/generics.go b/vendor/github.com/swaggo/swag/generics.go index 07344bba..80e93a90 100644 --- a/vendor/github.com/swaggo/swag/generics.go +++ b/vendor/github.com/swaggo/swag/generics.go @@ -61,6 +61,7 @@ func (pkgDefs *PackagesDefinitions) getTypeFromGenericParam(genericParam string, Enums: typeSpecDef.Enums, PkgPath: typeSpecDef.PkgPath, ParentSpec: typeSpecDef.ParentSpec, + SchemaName: "array_" + typeSpecDef.SchemaName, NotUnique: false, } } @@ -96,9 +97,9 @@ func (pkgDefs *PackagesDefinitions) getTypeFromGenericParam(genericParam string, Enums: typeSpecDef.Enums, PkgPath: typeSpecDef.PkgPath, ParentSpec: typeSpecDef.ParentSpec, + SchemaName: "map_" + parts[0] + "_" + typeSpecDef.SchemaName, NotUnique: false, } - } if IsGolangPrimitiveType(genericParam) { return &TypeSpecDef{ @@ -106,6 +107,7 @@ func (pkgDefs *PackagesDefinitions) getTypeFromGenericParam(genericParam string, Name: ast.NewIdent(genericParam), Type: ast.NewIdent(genericParam), }, + SchemaName: genericParam, } } return pkgDefs.FindTypeSpec(genericParam, file) @@ -155,14 +157,27 @@ func (pkgDefs *PackagesDefinitions) parametrizeGenericType(file *ast.File, origi } name = fmt.Sprintf("%s%s-", string(IgnoreNameOverridePrefix), original.TypeName()) + schemaName := fmt.Sprintf("%s-", original.SchemaName) + var nameParts []string + var schemaNameParts []string + for _, def := range formals { if specDef, ok := genericParamTypeDefs[def.Name]; ok { - nameParts = append(nameParts, specDef.TypeName()) + nameParts = append(nameParts, specDef.Name) + + schemaNamePart := specDef.Name + + if specDef.TypeSpec != nil { + schemaNamePart = specDef.TypeSpec.SchemaName + } + + schemaNameParts = append(schemaNameParts, schemaNamePart) } } name += normalizeGenericTypeName(strings.Join(nameParts, "-")) + schemaName += normalizeGenericTypeName(strings.Join(schemaNameParts, "-")) if typeSpec, ok := pkgDefs.uniqueDefinitions[name]; ok { return typeSpec @@ -180,6 +195,7 @@ func (pkgDefs *PackagesDefinitions) parametrizeGenericType(file *ast.File, origi Doc: original.TypeSpec.Doc, Assign: original.TypeSpec.Assign, }, + SchemaName: schemaName, } pkgDefs.uniqueDefinitions[name] = parametrizedTypeSpec diff --git a/vendor/github.com/swaggo/swag/operation.go b/vendor/github.com/swaggo/swag/operation.go index 169510ff..ba3a05b5 100644 --- a/vendor/github.com/swaggo/swag/operation.go +++ b/vendor/github.com/swaggo/swag/operation.go @@ -281,7 +281,7 @@ func (operation *Operation) ParseParamComment(commentLine string, astFile *ast.F requiredText := strings.ToLower(matches[4]) required := requiredText == "true" || requiredText == requiredLabel - description := matches[5] + description := strings.Join(strings.Split(matches[5], "\\n"), "\n") param := createParameter(paramType, description, name, objectType, refType, required, enums, operation.parser.collectionFormatInQuery) @@ -334,6 +334,9 @@ func (operation *Operation) ParseParamComment(commentLine string, astFile *ast.F if len(itemSchema.Type) == 0 { itemSchema = operation.parser.getUnderlyingSchema(prop.Items.Schema) } + if itemSchema == nil { + continue + } if len(itemSchema.Type) == 0 { continue } @@ -408,6 +411,7 @@ const ( exampleTag = "example" schemaExampleTag = "schemaExample" formatTag = "format" + titleTag = "title" validateTag = "validate" minimumTag = "minimum" maximumTag = "maximum" @@ -703,7 +707,7 @@ func parseMimeTypeList(mimeTypeList string, typeList *[]string, format string) e return nil } -var routerPattern = regexp.MustCompile(`^(/[\w./\-{}+:$]*)[[:blank:]]+\[(\w+)]`) +var routerPattern = regexp.MustCompile(`^(/[\w./\-{}\(\)+:$]*)[[:blank:]]+\[(\w+)]`) // ParseRouterComment parses comment for given `router` comment string. func (operation *Operation) ParseRouterComment(commentLine string, deprecated bool) error { @@ -837,9 +841,9 @@ func parseObjectSchema(parser *Parser, refType string, astFile *ast.File) (*spec case refType == NIL: return nil, nil case refType == INTERFACE: - return PrimitiveSchema(OBJECT), nil + return &spec.Schema{}, nil case refType == ANY: - return PrimitiveSchema(OBJECT), nil + return &spec.Schema{}, nil case IsGolangPrimitiveType(refType): refType = TransToValidSchemeType(refType) diff --git a/vendor/github.com/swaggo/swag/packages.go b/vendor/github.com/swaggo/swag/packages.go index 69a1b052..20324140 100644 --- a/vendor/github.com/swaggo/swag/packages.go +++ b/vendor/github.com/swaggo/swag/packages.go @@ -93,7 +93,7 @@ func (pkgDefs *PackagesDefinitions) RangeFiles(handle func(info *AstFileInfo) er for _, info := range pkgDefs.files { // ignore package path prefix with 'vendor' or $GOROOT, // because the router info of api will not be included these files. - if strings.HasPrefix(info.PackagePath, "vendor") || strings.HasPrefix(info.Path, runtime.GOROOT()) { + if strings.HasPrefix(info.PackagePath, "vendor") || (runtime.GOROOT() != "" && strings.HasPrefix(info.Path, runtime.GOROOT()+string(filepath.Separator))) { continue } sortedFiles = append(sortedFiles, info) @@ -166,6 +166,8 @@ func (pkgDefs *PackagesDefinitions) parseTypesFromFile(astFile *ast.File, packag pkgDefs.uniqueDefinitions[fullName] = nil anotherTypeDef.NotUnique = true pkgDefs.uniqueDefinitions[anotherTypeDef.TypeName()] = anotherTypeDef + anotherTypeDef.SetSchemaName() + typeSpecDef.NotUnique = true fullName = typeSpecDef.TypeName() pkgDefs.uniqueDefinitions[fullName] = typeSpecDef @@ -174,6 +176,8 @@ func (pkgDefs *PackagesDefinitions) parseTypesFromFile(astFile *ast.File, packag pkgDefs.uniqueDefinitions[fullName] = typeSpecDef } + typeSpecDef.SetSchemaName() + if pkgDefs.packages[typeSpecDef.PkgPath] == nil { pkgDefs.packages[typeSpecDef.PkgPath] = NewPackageDefinitions(astFile.Name.Name, typeSpecDef.PkgPath).AddTypeSpec(typeSpecDef.Name(), typeSpecDef) } else if _, ok = pkgDefs.packages[typeSpecDef.PkgPath].TypeDefinitions[typeSpecDef.Name()]; !ok { @@ -192,6 +196,7 @@ func (pkgDefs *PackagesDefinitions) parseFunctionScopedTypesFromFile(astFile *as for _, astDeclaration := range astFile.Decls { funcDeclaration, ok := astDeclaration.(*ast.FuncDecl) if ok && funcDeclaration.Body != nil { + functionScopedTypes := make(map[string]*TypeSpecDef) for _, stmt := range funcDeclaration.Body.List { if declStmt, ok := (stmt).(*ast.DeclStmt); ok { if genDecl, ok := (declStmt.Decl).(*ast.GenDecl); ok && genDecl.Tok == token.TYPE { @@ -212,12 +217,31 @@ func (pkgDefs *PackagesDefinitions) parseFunctionScopedTypesFromFile(astFile *as } } + fullName := typeSpecDef.TypeName() + if structType, ok := typeSpecDef.TypeSpec.Type.(*ast.StructType); ok { + for _, field := range structType.Fields.List { + var idt *ast.Ident + var ok bool + switch field.Type.(type) { + case *ast.Ident: + idt, ok = field.Type.(*ast.Ident) + case *ast.StarExpr: + idt, ok = field.Type.(*ast.StarExpr).X.(*ast.Ident) + case *ast.ArrayType: + idt, ok = field.Type.(*ast.ArrayType).Elt.(*ast.Ident) + } + if ok && !IsGolangPrimitiveType(idt.Name) { + if functype, ok := functionScopedTypes[idt.Name]; ok { + idt.Name = functype.TypeName() + } + } + } + } + if pkgDefs.uniqueDefinitions == nil { pkgDefs.uniqueDefinitions = make(map[string]*TypeSpecDef) } - fullName := typeSpecDef.TypeName() - anotherTypeDef, ok := pkgDefs.uniqueDefinitions[fullName] if ok { if anotherTypeDef == nil { @@ -228,14 +252,19 @@ func (pkgDefs *PackagesDefinitions) parseFunctionScopedTypesFromFile(astFile *as pkgDefs.uniqueDefinitions[fullName] = nil anotherTypeDef.NotUnique = true pkgDefs.uniqueDefinitions[anotherTypeDef.TypeName()] = anotherTypeDef + anotherTypeDef.SetSchemaName() + typeSpecDef.NotUnique = true fullName = typeSpecDef.TypeName() pkgDefs.uniqueDefinitions[fullName] = typeSpecDef } } else { pkgDefs.uniqueDefinitions[fullName] = typeSpecDef + functionScopedTypes[typeSpec.Name.Name] = typeSpecDef } + typeSpecDef.SetSchemaName() + if pkgDefs.packages[typeSpecDef.PkgPath] == nil { pkgDefs.packages[typeSpecDef.PkgPath] = NewPackageDefinitions(astFile.Name.Name, typeSpecDef.PkgPath).AddTypeSpec(fullName, typeSpecDef) } else if _, ok = pkgDefs.packages[typeSpecDef.PkgPath].TypeDefinitions[fullName]; !ok { @@ -353,7 +382,7 @@ func (pkgDefs *PackagesDefinitions) collectConstEnums(parsedSchemas map[*TypeSpe continue } - //delete it from parsed schemas, and will parse it again + // delete it from parsed schemas, and will parse it again if _, ok = parsedSchemas[typeDef]; ok { delete(parsedSchemas, typeDef) } @@ -469,7 +498,7 @@ func (pkgDefs *PackagesDefinitions) findPackagePathFromImports(pkg string, file } break } else if imp.Name.Name == "_" && len(pkg) > 0 { - //for unused types + // for unused types pd, ok := pkgDefs.packages[path] if ok { if pd.Name == pkg { @@ -558,17 +587,27 @@ func (pkgDefs *PackagesDefinitions) FindTypeSpec(typeName string, file *ast.File return typeDef } - //in case that comment //@name renamed the type with a name without a dot - typeDef, ok = pkgDefs.uniqueDefinitions[typeName] - if ok { - return typeDef - } - name := parts[0] typeDef, ok = pkgDefs.uniqueDefinitions[fullTypeName(file.Name.Name, name)] if !ok { pkgPaths, externalPkgPaths := pkgDefs.findPackagePathFromImports("", file) typeDef = pkgDefs.findTypeSpecFromPackagePaths(pkgPaths, externalPkgPaths, name) } - return pkgDefs.parametrizeGenericType(file, typeDef, typeName) + + if typeDef != nil { + return pkgDefs.parametrizeGenericType(file, typeDef, typeName) + } + + // in case that comment //@name renamed the type with a name without a dot + for k, v := range pkgDefs.uniqueDefinitions { + if v == nil { + pkgDefs.debug.Printf("%s TypeSpecDef is nil", k) + continue + } + if v.SchemaName == typeName { + return v + } + } + + return nil } diff --git a/vendor/github.com/swaggo/swag/parser.go b/vendor/github.com/swaggo/swag/parser.go index 604f8278..cd2a46c6 100644 --- a/vendor/github.com/swaggo/swag/parser.go +++ b/vendor/github.com/swaggo/swag/parser.go @@ -179,6 +179,9 @@ type Parser struct { // HostState is the state of the host HostState string + + // ParseFuncBody whether swag should parse api info inside of funcs + ParseFuncBody bool } // FieldParserFactory create FieldParser. @@ -187,7 +190,7 @@ type FieldParserFactory func(ps *Parser, field *ast.Field) FieldParser // FieldParser parse struct field. type FieldParser interface { ShouldSkip() bool - FieldName() (string, error) + FieldNames() ([]string, error) FormName() string HeaderName() string PathName() string @@ -511,7 +514,7 @@ func (parser *Parser) ParseGeneralAPIInfo(mainAPIFile string) error { func parseGeneralAPIInfo(parser *Parser, comments []string) error { previousAttribute := "" - + var tag *spec.Tag // parsing classic meta data model for line := 0; line < len(comments); line++ { commentLine := comments[line] @@ -572,42 +575,43 @@ func parseGeneralAPIInfo(parser *Parser, comments []string) error { case "@schemes": parser.swagger.Schemes = strings.Split(value, " ") case "@tag.name": - parser.swagger.Tags = append(parser.swagger.Tags, spec.Tag{ - TagProps: spec.TagProps{ - Name: value, - }, - }) + if parser.matchTag(value) { + parser.swagger.Tags = append(parser.swagger.Tags, spec.Tag{ + TagProps: spec.TagProps{ + Name: value, + }, + }) + tag = &parser.swagger.Tags[len(parser.swagger.Tags)-1] + } else { + tag = nil + } case "@tag.description": - tag := parser.swagger.Tags[len(parser.swagger.Tags)-1] - tag.TagProps.Description = value - replaceLastTag(parser.swagger.Tags, tag) + if tag != nil { + tag.TagProps.Description = value + } case "@tag.description.markdown": - tag := parser.swagger.Tags[len(parser.swagger.Tags)-1] + if tag != nil { + commentInfo, err := getMarkdownForTag(tag.TagProps.Name, parser.markdownFileDir) + if err != nil { + return err + } - commentInfo, err := getMarkdownForTag(tag.TagProps.Name, parser.markdownFileDir) - if err != nil { - return err + tag.TagProps.Description = string(commentInfo) } - - tag.TagProps.Description = string(commentInfo) - replaceLastTag(parser.swagger.Tags, tag) case "@tag.docs.url": - tag := parser.swagger.Tags[len(parser.swagger.Tags)-1] - tag.TagProps.ExternalDocs = &spec.ExternalDocumentation{ - URL: value, - Description: "", + if tag != nil { + tag.TagProps.ExternalDocs = &spec.ExternalDocumentation{ + URL: value, + } } - - replaceLastTag(parser.swagger.Tags, tag) case "@tag.docs.description": - tag := parser.swagger.Tags[len(parser.swagger.Tags)-1] - if tag.TagProps.ExternalDocs == nil { - return fmt.Errorf("%s needs to come after a @tags.docs.url", attribute) - } - - tag.TagProps.ExternalDocs.Description = value - replaceLastTag(parser.swagger.Tags, tag) + if tag != nil { + if tag.TagProps.ExternalDocs == nil { + return fmt.Errorf("%s needs to come after a @tags.docs.url", attribute) + } + tag.TagProps.ExternalDocs.Description = value + } case secBasicAttr, secAPIKeyAttr, secApplicationAttr, secImplicitAttr, secPasswordAttr, secAccessCodeAttr: scheme, err := parseSecAttributes(attribute, comments, &line) if err != nil { @@ -773,7 +777,10 @@ loopline: // Not mandatory field if securityAttr == descriptionAttr { - description = value + if description != "" { + description += "\n" + } + description += value } // next securityDefinitions @@ -881,6 +888,13 @@ func isGeneralAPIComment(comments []string) bool { } func getMarkdownForTag(tagName string, dirPath string) ([]byte, error) { + if tagName == "" { + // this happens when parsing the @description.markdown attribute + // it will be called properly another time with tagName="api" + // so we can safely return an empty byte slice here + return make([]byte, 0), nil + } + dirEntries, err := os.ReadDir(dirPath) if err != nil { return nil, err @@ -893,11 +907,12 @@ func getMarkdownForTag(tagName string, dirPath string) ([]byte, error) { fileName := entry.Name() - if !strings.Contains(fileName, ".md") { - continue + expectedFileName := tagName + if !strings.HasSuffix(tagName, ".md") { + expectedFileName = tagName + ".md" } - if strings.Contains(fileName, tagName) { + if fileName == expectedFileName { fullPath := filepath.Join(dirPath, fileName) commentInfo, err := os.ReadFile(fullPath) @@ -943,6 +958,27 @@ func getTagsFromComment(comment string) (tags []string) { } +func (parser *Parser) matchTag(tag string) bool { + if len(parser.tags) == 0 { + return true + } + + if _, has := parser.tags["!"+tag]; has { + return false + } + if _, has := parser.tags[tag]; has { + return true + } + + // If all tags are negation then we should return true + for key := range parser.tags { + if key[0] != '!' { + return false + } + } + return true +} + func (parser *Parser) matchTags(comments []*ast.Comment) (match bool) { if len(parser.tags) == 0 { return true @@ -991,30 +1027,28 @@ func matchExtension(extensionToMatch string, comments []*ast.Comment) (match boo // ParseRouterAPIInfo parses router api info for given astFile. func (parser *Parser) ParseRouterAPIInfo(fileInfo *AstFileInfo) error { -DeclsLoop: - for _, astDescription := range fileInfo.File.Decls { - if (fileInfo.ParseFlag & ParseOperations) == ParseNone { - continue + if (fileInfo.ParseFlag & ParseOperations) == ParseNone { + return nil + } + + // parse File.Comments instead of File.Decls.Doc if ParseFuncBody flag set to "true" + if parser.ParseFuncBody { + for _, astComments := range fileInfo.File.Comments { + if astComments.List != nil { + if err := parser.parseRouterAPIInfoComment(astComments.List, fileInfo); err != nil { + return err + } + } } + + return nil + } + + for _, astDescription := range fileInfo.File.Decls { astDeclaration, ok := astDescription.(*ast.FuncDecl) if ok && astDeclaration.Doc != nil && astDeclaration.Doc.List != nil { - if parser.matchTags(astDeclaration.Doc.List) && - matchExtension(parser.parseExtension, astDeclaration.Doc.List) { - // for per 'function' comment, create a new 'Operation' object - operation := NewOperation(parser, SetCodeExampleFilesDirectory(parser.codeExampleFilesDir)) - for _, comment := range astDeclaration.Doc.List { - err := operation.ParseComment(comment.Text, fileInfo.File) - if err != nil { - return fmt.Errorf("ParseComment error in file %s :%+v", fileInfo.Path, err) - } - if operation.State != "" && operation.State != parser.HostState { - continue DeclsLoop - } - } - err := processRouterOperation(parser, operation) - if err != nil { - return err - } + if err := parser.parseRouterAPIInfoComment(astDeclaration.Doc.List, fileInfo); err != nil { + return err } } } @@ -1022,6 +1056,28 @@ DeclsLoop: return nil } +func (parser *Parser) parseRouterAPIInfoComment(comments []*ast.Comment, fileInfo *AstFileInfo) error { + if parser.matchTags(comments) && matchExtension(parser.parseExtension, comments) { + // for per 'function' comment, create a new 'Operation' object + operation := NewOperation(parser, SetCodeExampleFilesDirectory(parser.codeExampleFilesDir)) + for _, comment := range comments { + err := operation.ParseComment(comment.Text, fileInfo.File) + if err != nil { + return fmt.Errorf("ParseComment error in file %s :%+v", fileInfo.Path, err) + } + if operation.State != "" && operation.State != parser.HostState { + return nil + } + } + err := processRouterOperation(parser, operation) + if err != nil { + return err + } + } + + return nil +} + func refRouteMethodOp(item *spec.PathItem, method string) (op **spec.Operation) { switch method { case http.MethodGet: @@ -1238,7 +1294,10 @@ func (parser *Parser) ParseDefinition(typeSpecDef *TypeSpecDef) (*Schema, error) } if definition.Description == "" { - fillDefinitionDescription(definition, typeSpecDef.File, typeSpecDef) + err = parser.fillDefinitionDescription(definition, typeSpecDef.File, typeSpecDef) + if err != nil { + return nil, err + } } if len(typeSpecDef.Enums) > 0 { @@ -1260,8 +1319,14 @@ func (parser *Parser) ParseDefinition(typeSpecDef *TypeSpecDef) (*Schema, error) } } + schemaName := typeName + + if typeSpecDef.SchemaName != "" { + schemaName = typeSpecDef.SchemaName + } + sch := Schema{ - Name: typeName, + Name: schemaName, PkgPath: typeSpecDef.PkgPath, Schema: definition, } @@ -1282,7 +1347,7 @@ func fullTypeName(parts ...string) string { // fillDefinitionDescription additionally fills fields in definition (spec.Schema) // TODO: If .go file contains many types, it may work for a long time -func fillDefinitionDescription(definition *spec.Schema, file *ast.File, typeSpecDef *TypeSpecDef) { +func (parser *Parser) fillDefinitionDescription(definition *spec.Schema, file *ast.File, typeSpecDef *TypeSpecDef) (err error) { if file == nil { return } @@ -1297,16 +1362,23 @@ func fillDefinitionDescription(definition *spec.Schema, file *ast.File, typeSpec if !ok || typeSpec != typeSpecDef.TypeSpec { continue } - - definition.Description = - extractDeclarationDescription(typeSpec.Doc, typeSpec.Comment, generalDeclaration.Doc) + var typeName string + if typeSpec.Name != nil { + typeName = typeSpec.Name.Name + } + definition.Description, err = + parser.extractDeclarationDescription(typeName, typeSpec.Doc, typeSpec.Comment, generalDeclaration.Doc) + if err != nil { + return + } } } + return nil } // extractDeclarationDescription gets first description // from attribute descriptionAttr in commentGroups (ast.CommentGroup) -func extractDeclarationDescription(commentGroups ...*ast.CommentGroup) string { +func (parser *Parser) extractDeclarationDescription(typeName string, commentGroups ...*ast.CommentGroup) (string, error) { var description string for _, commentGroup := range commentGroups { @@ -1321,9 +1393,23 @@ func extractDeclarationDescription(commentGroups ...*ast.CommentGroup) string { if len(commentText) == 0 { continue } - attribute := FieldsByAnySpace(commentText, 2)[0] + fields := FieldsByAnySpace(commentText, 2) + attribute := fields[0] - if strings.ToLower(attribute) != descriptionAttr { + if attr := strings.ToLower(attribute); attr == descriptionMarkdownAttr { + if len(fields) > 1 { + typeName = fields[1] + } + if typeName == "" { + continue + } + desc, err := getMarkdownForTag(typeName, parser.markdownFileDir) + if err != nil { + return "", err + } + // if found markdown description, we will only use the markdown file content + return string(desc), nil + } else if attr != descriptionAttr { if !isHandlingDescription { continue } @@ -1336,7 +1422,7 @@ func extractDeclarationDescription(commentGroups ...*ast.CommentGroup) string { } } - return strings.TrimLeft(description, " ") + return strings.TrimLeft(description, " "), nil } // parseTypeExpr parses given type expression that corresponds to the type under @@ -1441,20 +1527,20 @@ func (parser *Parser) parseStructField(file *ast.File, field *ast.Field) (map[st return nil, nil, nil } - fieldName, err := ps.FieldName() + fieldNames, err := ps.FieldNames() if err != nil { return nil, nil, err } - if fieldName == "" { + if len(fieldNames) == 0 { typeName, err := getFieldType(file, field.Type, nil) if err != nil { - return nil, nil, fmt.Errorf("%s: %w", fieldName, err) + return nil, nil, err } schema, err := parser.getTypeSchema(typeName, file, false) if err != nil { - return nil, nil, fmt.Errorf("%s: %w", fieldName, err) + return nil, nil, err } if len(schema.Type) > 0 && schema.Type[0] == OBJECT { @@ -1476,7 +1562,7 @@ func (parser *Parser) parseStructField(file *ast.File, field *ast.Field) (map[st schema, err := ps.CustomSchema() if err != nil { - return nil, nil, fmt.Errorf("%s: %w", fieldName, err) + return nil, nil, fmt.Errorf("%v: %w", fieldNames, err) } if schema == nil { @@ -1490,24 +1576,24 @@ func (parser *Parser) parseStructField(file *ast.File, field *ast.Field) (map[st } if err != nil { - return nil, nil, fmt.Errorf("%s: %w", fieldName, err) + return nil, nil, fmt.Errorf("%v: %w", fieldNames, err) } } err = ps.ComplementSchema(schema) if err != nil { - return nil, nil, fmt.Errorf("%s: %w", fieldName, err) + return nil, nil, fmt.Errorf("%v: %w", fieldNames, err) } var tagRequired []string required, err := ps.IsRequired() if err != nil { - return nil, nil, fmt.Errorf("%s: %w", fieldName, err) + return nil, nil, fmt.Errorf("%v: %w", fieldNames, err) } if required { - tagRequired = append(tagRequired, fieldName) + tagRequired = append(tagRequired, fieldNames...) } if schema.Extensions == nil { @@ -1522,8 +1608,11 @@ func (parser *Parser) parseStructField(file *ast.File, field *ast.Field) (map[st if pathName := ps.PathName(); len(pathName) > 0 { schema.Extensions["path"] = pathName } - - return map[string]spec.Schema{fieldName: *schema}, tagRequired, nil + fields := make(map[string]spec.Schema) + for _, name := range fieldNames { + fields[name] = *schema + } + return fields, tagRequired, nil } func getFieldType(file *ast.File, field ast.Expr, genericParamTypeDefs map[string]*genericTypeSpec) (string, error) { @@ -1609,10 +1698,6 @@ func (parser *Parser) GetSchemaTypePath(schema *spec.Schema, depth int) []string return []string{ANY} } -func replaceLastTag(slice []spec.Tag, element spec.Tag) { - slice = append(slice[:len(slice)-1], element) -} - // defineTypeOfExample example value define the type (object and array unsupported). func defineTypeOfExample(schemaType, arrayType, exampleValue string) (interface{}, error) { switch schemaType { diff --git a/vendor/github.com/swaggo/swag/types.go b/vendor/github.com/swaggo/swag/types.go index 0076a6b4..5f3031e0 100644 --- a/vendor/github.com/swaggo/swag/types.go +++ b/vendor/github.com/swaggo/swag/types.go @@ -30,6 +30,8 @@ type TypeSpecDef struct { PkgPath string ParentSpec ast.Decl + SchemaName string + NotUnique bool } @@ -46,20 +48,6 @@ func (t *TypeSpecDef) Name() string { func (t *TypeSpecDef) TypeName() string { if ignoreNameOverride(t.TypeSpec.Name.Name) { return t.TypeSpec.Name.Name[1:] - } else if t.TypeSpec.Comment != nil { - // get alias from comment '// @name ' - const regexCaseInsensitive = "(?i)" - reTypeName, err := regexp.Compile(regexCaseInsensitive + `^@name\s+(\S+)`) - if err != nil { - panic(err) - } - for _, comment := range t.TypeSpec.Comment.List { - trimmedComment := strings.TrimSpace(strings.TrimLeft(comment.Text, "/")) - texts := reTypeName.FindStringSubmatch(trimmedComment) - if len(texts) > 1 { - return texts[1] - } - } } var names []string @@ -86,6 +74,36 @@ func (t *TypeSpecDef) FullPath() string { return t.PkgPath + "." + t.Name() } +const regexCaseInsensitive = "(?i)" + +var reTypeName = regexp.MustCompile(regexCaseInsensitive + `^@name\s+(\S+)`) + +func (t *TypeSpecDef) Alias() string { + if t.TypeSpec.Comment == nil { + return "" + } + + // get alias from comment '// @name ' + for _, comment := range t.TypeSpec.Comment.List { + trimmedComment := strings.TrimSpace(strings.TrimLeft(comment.Text, "/")) + texts := reTypeName.FindStringSubmatch(trimmedComment) + if len(texts) > 1 { + return texts[1] + } + } + + return "" +} + +func (t *TypeSpecDef) SetSchemaName() { + if alias := t.Alias(); alias != "" { + t.SchemaName = alias + return + } + + t.SchemaName = t.TypeName() +} + // AstFileInfo information of an ast.File. type AstFileInfo struct { //FileSet the FileSet object which is used to parse this go source file diff --git a/vendor/github.com/swaggo/swag/version.go b/vendor/github.com/swaggo/swag/version.go index ff2810e6..cf52320f 100644 --- a/vendor/github.com/swaggo/swag/version.go +++ b/vendor/github.com/swaggo/swag/version.go @@ -1,4 +1,4 @@ package swag // Version of swag. -const Version = "v1.16.3" +const Version = "v1.16.4" diff --git a/vendor/github.com/vektah/gqlparser/v2/validator/rules/fields_on_correct_type.go b/vendor/github.com/vektah/gqlparser/v2/validator/rules/fields_on_correct_type.go index daa86448..64392185 100644 --- a/vendor/github.com/vektah/gqlparser/v2/validator/rules/fields_on_correct_type.go +++ b/vendor/github.com/vektah/gqlparser/v2/validator/rules/fields_on_correct_type.go @@ -11,27 +11,40 @@ import ( . "github.com/vektah/gqlparser/v2/validator" ) -var FieldsOnCorrectTypeRule = Rule{ - Name: "FieldsOnCorrectType", - RuleFunc: func(observers *Events, addError AddErrFunc) { - observers.OnField(func(walker *Walker, field *ast.Field) { - if field.ObjectDefinition == nil || field.Definition != nil { - return - } +func ruleFuncFieldsOnCorrectType(observers *Events, addError AddErrFunc, disableSuggestion bool) { + observers.OnField(func(walker *Walker, field *ast.Field) { + if field.ObjectDefinition == nil || field.Definition != nil { + return + } - message := fmt.Sprintf(`Cannot query field "%s" on type "%s".`, field.Name, field.ObjectDefinition.Name) + message := fmt.Sprintf(`Cannot query field "%s" on type "%s".`, field.Name, field.ObjectDefinition.Name) + if !disableSuggestion { if suggestedTypeNames := getSuggestedTypeNames(walker, field.ObjectDefinition, field.Name); suggestedTypeNames != nil { message += " Did you mean to use an inline fragment on " + QuotedOrList(suggestedTypeNames...) + "?" } else if suggestedFieldNames := getSuggestedFieldNames(field.ObjectDefinition, field.Name); suggestedFieldNames != nil { message += " Did you mean " + QuotedOrList(suggestedFieldNames...) + "?" } + } + + addError( + Message("%s", message), + At(field.Position), + ) + }) +} + +var FieldsOnCorrectTypeRule = Rule{ + Name: "FieldsOnCorrectType", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncFieldsOnCorrectType(observers, addError, false) + }, +} - addError( - Message("%s", message), - At(field.Position), - ) - }) +var FieldsOnCorrectTypeRuleWithoutSuggestions = Rule{ + Name: "FieldsOnCorrectTypeWithoutSuggestions", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncFieldsOnCorrectType(observers, addError, true) }, } diff --git a/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_argument_names.go b/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_argument_names.go index 2659aeb5..dbb38a32 100644 --- a/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_argument_names.go +++ b/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_argument_names.go @@ -7,43 +7,53 @@ import ( . "github.com/vektah/gqlparser/v2/validator" ) -var KnownArgumentNamesRule = Rule{ - Name: "KnownArgumentNames", - RuleFunc: func(observers *Events, addError AddErrFunc) { - // A GraphQL field is only valid if all supplied arguments are defined by that field. - observers.OnField(func(walker *Walker, field *ast.Field) { - if field.Definition == nil || field.ObjectDefinition == nil { - return +func ruleFuncKnownArgumentNames(observers *Events, addError AddErrFunc, disableSuggestion bool) { + // A GraphQL field is only valid if all supplied arguments are defined by that field. + observers.OnField(func(walker *Walker, field *ast.Field) { + if field.Definition == nil || field.ObjectDefinition == nil { + return + } + for _, arg := range field.Arguments { + def := field.Definition.Arguments.ForName(arg.Name) + if def != nil { + continue } - for _, arg := range field.Arguments { - def := field.Definition.Arguments.ForName(arg.Name) - if def != nil { - continue - } + if disableSuggestion { + addError( + Message(`Unknown argument "%s" on field "%s.%s".`, arg.Name, field.ObjectDefinition.Name, field.Name), + At(field.Position), + ) + } else { var suggestions []string for _, argDef := range field.Definition.Arguments { suggestions = append(suggestions, argDef.Name) } - addError( Message(`Unknown argument "%s" on field "%s.%s".`, arg.Name, field.ObjectDefinition.Name, field.Name), SuggestListQuoted("Did you mean", arg.Name, suggestions), At(field.Position), ) } - }) + } + }) - observers.OnDirective(func(walker *Walker, directive *ast.Directive) { - if directive.Definition == nil { - return + observers.OnDirective(func(walker *Walker, directive *ast.Directive) { + if directive.Definition == nil { + return + } + for _, arg := range directive.Arguments { + def := directive.Definition.Arguments.ForName(arg.Name) + if def != nil { + continue } - for _, arg := range directive.Arguments { - def := directive.Definition.Arguments.ForName(arg.Name) - if def != nil { - continue - } + if disableSuggestion { + addError( + Message(`Unknown argument "%s" on directive "@%s".`, arg.Name, directive.Name), + At(directive.Position), + ) + } else { var suggestions []string for _, argDef := range directive.Definition.Arguments { suggestions = append(suggestions, argDef.Name) @@ -55,7 +65,21 @@ var KnownArgumentNamesRule = Rule{ At(directive.Position), ) } - }) + } + }) +} + +var KnownArgumentNamesRule = Rule{ + Name: "KnownArgumentNames", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncKnownArgumentNames(observers, addError, false) + }, +} + +var KnownArgumentNamesRuleWithoutSuggestions = Rule{ + Name: "KnownArgumentNamesWithoutSuggestions", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncKnownArgumentNames(observers, addError, true) }, } diff --git a/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_type_names.go b/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_type_names.go index 717019fb..75a0e861 100644 --- a/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_type_names.go +++ b/vendor/github.com/vektah/gqlparser/v2/validator/rules/known_type_names.go @@ -7,46 +7,50 @@ import ( . "github.com/vektah/gqlparser/v2/validator" ) -var KnownTypeNamesRule = Rule{ - Name: "KnownTypeNames", - RuleFunc: func(observers *Events, addError AddErrFunc) { - observers.OnVariable(func(walker *Walker, variable *ast.VariableDefinition) { - typeName := variable.Type.Name() - typdef := walker.Schema.Types[typeName] - if typdef != nil { - return - } +func ruleFuncKnownTypeNames(observers *Events, addError AddErrFunc, disableSuggestion bool) { + observers.OnVariable(func(walker *Walker, variable *ast.VariableDefinition) { + typeName := variable.Type.Name() + typdef := walker.Schema.Types[typeName] + if typdef != nil { + return + } - addError( - Message(`Unknown type "%s".`, typeName), - At(variable.Position), - ) - }) + addError( + Message(`Unknown type "%s".`, typeName), + At(variable.Position), + ) + }) - observers.OnInlineFragment(func(walker *Walker, inlineFragment *ast.InlineFragment) { - typedName := inlineFragment.TypeCondition - if typedName == "" { - return - } + observers.OnInlineFragment(func(walker *Walker, inlineFragment *ast.InlineFragment) { + typedName := inlineFragment.TypeCondition + if typedName == "" { + return + } - def := walker.Schema.Types[typedName] - if def != nil { - return - } + def := walker.Schema.Types[typedName] + if def != nil { + return + } - addError( - Message(`Unknown type "%s".`, typedName), - At(inlineFragment.Position), - ) - }) + addError( + Message(`Unknown type "%s".`, typedName), + At(inlineFragment.Position), + ) + }) - observers.OnFragment(func(walker *Walker, fragment *ast.FragmentDefinition) { - typeName := fragment.TypeCondition - def := walker.Schema.Types[typeName] - if def != nil { - return - } + observers.OnFragment(func(walker *Walker, fragment *ast.FragmentDefinition) { + typeName := fragment.TypeCondition + def := walker.Schema.Types[typeName] + if def != nil { + return + } + if disableSuggestion { + addError( + Message(`Unknown type "%s".`, typeName), + At(fragment.Position), + ) + } else { var possibleTypes []string for _, t := range walker.Schema.Types { possibleTypes = append(possibleTypes, t.Name) @@ -57,7 +61,21 @@ var KnownTypeNamesRule = Rule{ SuggestListQuoted("Did you mean", typeName, possibleTypes), At(fragment.Position), ) - }) + } + }) +} + +var KnownTypeNamesRule = Rule{ + Name: "KnownTypeNames", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncKnownTypeNames(observers, addError, false) + }, +} + +var KnownTypeNamesRuleWithoutSuggestions = Rule{ + Name: "KnownTypeNamesWithoutSuggestions", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncKnownTypeNames(observers, addError, true) }, } diff --git a/vendor/github.com/vektah/gqlparser/v2/validator/rules/values_of_correct_type.go b/vendor/github.com/vektah/gqlparser/v2/validator/rules/values_of_correct_type.go index 7784fe0c..6cfdc3dc 100644 --- a/vendor/github.com/vektah/gqlparser/v2/validator/rules/values_of_correct_type.go +++ b/vendor/github.com/vektah/gqlparser/v2/validator/rules/values_of_correct_type.go @@ -11,80 +11,98 @@ import ( . "github.com/vektah/gqlparser/v2/validator" ) -var ValuesOfCorrectTypeRule = Rule{ - Name: "ValuesOfCorrectType", - RuleFunc: func(observers *Events, addError AddErrFunc) { - observers.OnValue(func(walker *Walker, value *ast.Value) { - if value.Definition == nil || value.ExpectedType == nil { +func ruleFuncValuesOfCorrectType(observers *Events, addError AddErrFunc, disableSuggestion bool) { + observers.OnValue(func(walker *Walker, value *ast.Value) { + if value.Definition == nil || value.ExpectedType == nil { + return + } + + if value.Kind == ast.NullValue && value.ExpectedType.NonNull { + addError( + Message(`Expected value of type "%s", found %s.`, value.ExpectedType.String(), value.String()), + At(value.Position), + ) + } + + if value.Definition.Kind == ast.Scalar { + // Skip custom validating scalars + if !value.Definition.OneOf("Int", "Float", "String", "Boolean", "ID") { return } + } - if value.Kind == ast.NullValue && value.ExpectedType.NonNull { - addError( - Message(`Expected value of type "%s", found %s.`, value.ExpectedType.String(), value.String()), - At(value.Position), - ) + var possibleEnums []string + if value.Definition.Kind == ast.Enum { + for _, val := range value.Definition.EnumValues { + possibleEnums = append(possibleEnums, val.Name) } + } - if value.Definition.Kind == ast.Scalar { - // Skip custom validating scalars - if !value.Definition.OneOf("Int", "Float", "String", "Boolean", "ID") { - return - } - } + rawVal, err := value.Value(nil) + if err != nil { + unexpectedTypeMessage(addError, value) + } - var possibleEnums []string - if value.Definition.Kind == ast.Enum { - for _, val := range value.Definition.EnumValues { - possibleEnums = append(possibleEnums, val.Name) - } + switch value.Kind { + case ast.NullValue: + return + case ast.ListValue: + if value.ExpectedType.Elem == nil { + unexpectedTypeMessage(addError, value) + return } - rawVal, err := value.Value(nil) - if err != nil { + case ast.IntValue: + if !value.Definition.OneOf("Int", "Float", "ID") { unexpectedTypeMessage(addError, value) } - switch value.Kind { - case ast.NullValue: - return - case ast.ListValue: - if value.ExpectedType.Elem == nil { - unexpectedTypeMessage(addError, value) - return - } - - case ast.IntValue: - if !value.Definition.OneOf("Int", "Float", "ID") { - unexpectedTypeMessage(addError, value) - } - - case ast.FloatValue: - if !value.Definition.OneOf("Float") { - unexpectedTypeMessage(addError, value) - } + case ast.FloatValue: + if !value.Definition.OneOf("Float") { + unexpectedTypeMessage(addError, value) + } - case ast.StringValue, ast.BlockValue: - if value.Definition.Kind == ast.Enum { + case ast.StringValue, ast.BlockValue: + if value.Definition.Kind == ast.Enum { + if disableSuggestion { + addError( + Message(`Enum "%s" cannot represent non-enum value: %s.`, value.ExpectedType.String(), value.String()), + At(value.Position), + ) + } else { rawValStr := fmt.Sprint(rawVal) addError( Message(`Enum "%s" cannot represent non-enum value: %s.`, value.ExpectedType.String(), value.String()), SuggestListQuoted("Did you mean the enum value", rawValStr, possibleEnums), At(value.Position), ) - } else if !value.Definition.OneOf("String", "ID") { - unexpectedTypeMessage(addError, value) } + } else if !value.Definition.OneOf("String", "ID") { + unexpectedTypeMessage(addError, value) + } - case ast.EnumValue: - if value.Definition.Kind != ast.Enum { + case ast.EnumValue: + if value.Definition.Kind != ast.Enum { + if disableSuggestion { + addError( + unexpectedTypeMessageOnly(value), + At(value.Position), + ) + } else { rawValStr := fmt.Sprint(rawVal) addError( unexpectedTypeMessageOnly(value), SuggestListUnquoted("Did you mean the enum value", rawValStr, possibleEnums), At(value.Position), ) - } else if value.Definition.EnumValues.ForName(value.Raw) == nil { + } + } else if value.Definition.EnumValues.ForName(value.Raw) == nil { + if disableSuggestion { + addError( + Message(`Value "%s" does not exist in "%s" enum.`, value.String(), value.ExpectedType.String()), + At(value.Position), + ) + } else { rawValStr := fmt.Sprint(rawVal) addError( Message(`Value "%s" does not exist in "%s" enum.`, value.String(), value.ExpectedType.String()), @@ -92,29 +110,36 @@ var ValuesOfCorrectTypeRule = Rule{ At(value.Position), ) } + } - case ast.BooleanValue: - if !value.Definition.OneOf("Boolean") { - unexpectedTypeMessage(addError, value) - } + case ast.BooleanValue: + if !value.Definition.OneOf("Boolean") { + unexpectedTypeMessage(addError, value) + } - case ast.ObjectValue: - - for _, field := range value.Definition.Fields { - if field.Type.NonNull { - fieldValue := value.Children.ForName(field.Name) - if fieldValue == nil && field.DefaultValue == nil { - addError( - Message(`Field "%s.%s" of required type "%s" was not provided.`, value.Definition.Name, field.Name, field.Type.String()), - At(value.Position), - ) - continue - } + case ast.ObjectValue: + + for _, field := range value.Definition.Fields { + if field.Type.NonNull { + fieldValue := value.Children.ForName(field.Name) + if fieldValue == nil && field.DefaultValue == nil { + addError( + Message(`Field "%s.%s" of required type "%s" was not provided.`, value.Definition.Name, field.Name, field.Type.String()), + At(value.Position), + ) + continue } } + } - for _, fieldValue := range value.Children { - if value.Definition.Fields.ForName(fieldValue.Name) == nil { + for _, fieldValue := range value.Children { + if value.Definition.Fields.ForName(fieldValue.Name) == nil { + if disableSuggestion { + addError( + Message(`Field "%s" is not defined by type "%s".`, fieldValue.Name, value.Definition.Name), + At(fieldValue.Position), + ) + } else { var suggestions []string for _, fieldValue := range value.Definition.Fields { suggestions = append(suggestions, fieldValue.Name) @@ -127,14 +152,28 @@ var ValuesOfCorrectTypeRule = Rule{ ) } } + } - case ast.Variable: - return + case ast.Variable: + return - default: - panic(fmt.Errorf("unhandled %T", value)) - } - }) + default: + panic(fmt.Errorf("unhandled %T", value)) + } + }) +} + +var ValuesOfCorrectTypeRule = Rule{ + Name: "ValuesOfCorrectType", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncValuesOfCorrectType(observers, addError, false) + }, +} + +var ValuesOfCorrectTypeRuleWithoutSuggestions = Rule{ + Name: "ValuesOfCorrectTypeWithoutSuggestions", + RuleFunc: func(observers *Events, addError AddErrFunc) { + ruleFuncValuesOfCorrectType(observers, addError, true) }, } diff --git a/vendor/github.com/vektah/gqlparser/v2/validator/validator.go b/vendor/github.com/vektah/gqlparser/v2/validator/validator.go index 36564b23..fb4db194 100644 --- a/vendor/github.com/vektah/gqlparser/v2/validator/validator.go +++ b/vendor/github.com/vektah/gqlparser/v2/validator/validator.go @@ -23,6 +23,17 @@ func AddRule(name string, ruleFunc RuleFunc) { specifiedRules = append(specifiedRules, Rule{Name: name, RuleFunc: ruleFunc}) } +func RemoveRule(name string) { + var result []Rule // nolint:prealloc // using initialized with len(rules) produces a race condition + for _, r := range specifiedRules { + if r.Name == name { + continue + } + result = append(result, r) + } + specifiedRules = result +} + func Validate(schema *Schema, doc *QueryDocument, rules ...Rule) gqlerror.List { if rules == nil { rules = specifiedRules diff --git a/vendor/modules.txt b/vendor/modules.txt index 792d3807..115e8935 100644 --- a/vendor/modules.txt +++ b/vendor/modules.txt @@ -30,13 +30,13 @@ github.com/KyleBanks/depth # github.com/Masterminds/semver/v3 v3.3.0 ## explicit; go 1.21 github.com/Masterminds/semver/v3 -# github.com/adhocore/gronx v1.19.1 +# github.com/adhocore/gronx v1.19.3 ## explicit; go 1.13 github.com/adhocore/gronx # github.com/agnivade/levenshtein v1.2.0 ## explicit; go 1.21 github.com/agnivade/levenshtein -# github.com/andybalholm/brotli v1.1.0 +# github.com/andybalholm/brotli v1.1.1 ## explicit; go 1.13 github.com/andybalholm/brotli github.com/andybalholm/brotli/matchfinder @@ -117,13 +117,13 @@ github.com/dolthub/swiss/simd # github.com/dustin/go-humanize v1.0.1 ## explicit; go 1.16 github.com/dustin/go-humanize -# github.com/fatih/color v1.17.0 +# github.com/fatih/color v1.18.0 ## explicit; go 1.17 github.com/fatih/color # github.com/fujiwara/shapeio v1.0.0 ## explicit; go 1.16 github.com/fujiwara/shapeio -# github.com/gabriel-vasile/mimetype v1.4.5 +# github.com/gabriel-vasile/mimetype v1.4.6 ## explicit; go 1.20 github.com/gabriel-vasile/mimetype github.com/gabriel-vasile/mimetype/internal/charset @@ -239,7 +239,7 @@ github.com/joho/godotenv/autoload # github.com/josharian/intern v1.0.0 ## explicit; go 1.5 github.com/josharian/intern -# github.com/klauspost/compress v1.17.10 +# github.com/klauspost/compress v1.17.11 ## explicit; go 1.21 github.com/klauspost/compress github.com/klauspost/compress/flate @@ -301,8 +301,8 @@ github.com/miekg/dns # github.com/minio/md5-simd v1.1.2 ## explicit; go 1.14 github.com/minio/md5-simd -# github.com/minio/minio-go/v7 v7.0.77 -## explicit; go 1.21 +# github.com/minio/minio-go/v7 v7.0.80 +## explicit; go 1.22 github.com/minio/minio-go/v7 github.com/minio/minio-go/v7/pkg/cors github.com/minio/minio-go/v7/pkg/credentials @@ -327,7 +327,7 @@ github.com/pmezard/go-difflib/difflib # github.com/power-devops/perfstat v0.0.0-20240221224432-82ca36839d55 ## explicit; go 1.14 github.com/power-devops/perfstat -# github.com/prometheus/client_golang v1.20.4 +# github.com/prometheus/client_golang v1.20.5 ## explicit; go 1.20 github.com/prometheus/client_golang/internal/github.com/golang/gddo/httputil github.com/prometheus/client_golang/internal/github.com/golang/gddo/httputil/header @@ -337,7 +337,7 @@ github.com/prometheus/client_golang/prometheus/promhttp # github.com/prometheus/client_model v0.6.1 ## explicit; go 1.19 github.com/prometheus/client_model/go -# github.com/prometheus/common v0.60.0 +# github.com/prometheus/common v0.60.1 ## explicit; go 1.21 github.com/prometheus/common/expfmt github.com/prometheus/common/model @@ -380,7 +380,7 @@ github.com/swaggo/echo-swagger # github.com/swaggo/files/v2 v2.0.1 ## explicit; go 1.16 github.com/swaggo/files/v2 -# github.com/swaggo/swag v1.16.3 +# github.com/swaggo/swag v1.16.4 ## explicit; go 1.18 github.com/swaggo/swag # github.com/tklauser/go-sysconf v0.3.14 @@ -398,8 +398,8 @@ github.com/valyala/bytebufferpool # github.com/valyala/fasttemplate v1.2.2 ## explicit; go 1.12 github.com/valyala/fasttemplate -# github.com/vektah/gqlparser/v2 v2.5.17 -## explicit; go 1.19 +# github.com/vektah/gqlparser/v2 v2.5.18 +## explicit; go 1.21 github.com/vektah/gqlparser/v2 github.com/vektah/gqlparser/v2/ast github.com/vektah/gqlparser/v2/gqlerror